Taylor MacDonald: What's going on with your mustache? Are you are you going for the handlebars again? Like is this a is ⁓ Yeah. Is this like a trajectory towards your next birthday party or what?

Sean C Davis: Yeah, my my browser's been weird too. Yes, it's a that'd be that would be quite the investment 'cause that would be another ⁓ year and nine months. No, I ⁓ a change. ⁓ says we'll call it a summertime. Summertime stash. I think my my my eight year old

Taylor MacDonald: A stash. ⁓ It's summer sausage, but it's on your face. ⁓

Sean C Davis: Yeah, yeah, precisely. Doesn't taste as good either. ⁓ yeah. I one of one of my kids was was joking with me, like you should you should get a mustache and then the other one was pulling on my beard too frequently and I got annoyed and I also like to I generally act on dares. So here we are.

Taylor MacDonald: ⁓ man, that's funny. Yeah. ⁓ yeah. Who dared you? Was it the kids that dared you or your or your wife or who?

Sean C Davis: It was it was it was the kids. Yeah. And and I knew Katie would would be she'd she would it would be more of like a ⁓ eyebrow raise than a, you know, actually upset about it, but I intentionally did it on Father's Day and ⁓ ⁓ she was like, ⁓ you look different. And I was like, Yeah, I know, you can't say anything today. She's right. She's like, Yeah, I know, I'll tell you look stupid tomorrow ⁓

Taylor MacDonald: You just can't say anything. She's so funny. Alright, okay, okay. Let's back to back to work here. So, welcome back, Blood, Sweat and Tokens. This is episode eight. ⁓ had a really interesting conversation last last time, Sean. Did you happen to reflect on any of those conversations? We talked about the end of humanity. ⁓ We about the delegation of human intellect ⁓ at the consequence of brain atrophy. ⁓ we about All sorts of stuff. We talked about custom AI devices, we talked about capitalism. Needless to say, I've been reflecting on the conversation myself, and I thought was great. You talked about Cincy AI?

Sean C Davis: Yes. Yeah, I've I've been all over the place and I actually ⁓ so I I have been I've been so obsessively building over the last three or four months that I stopped I I stopped finding time in my day to read and so most of what I've been consuming has been listening. I do a I have a bicycle commute every day and so I've been popping in an ebook and I switched books this week and I And so just yesterday I started this new one. It is called it's called Against the Grain. And I don't know Well, there's okay, so there's there's actually the the book I just finished is called shoot, what was it? I think it's called The Righteous Mind. And it's we this this would take us way off topic and way deep down the rabbit hole like we did last week. But it was kind of it was this deep dive into trying to understand how people work and and trying to like push back against good and evil and just understand why. But this Against the Grain book, I don't know what's going what it's gonna what the argument is going to be, but I am sensing. So it's mostly taking place, mostly studying the Mesopotamia ancient world. And I think what it's going to start to try to refute is that the downfall of or like the it it I I have a sense that it's going to connect climate change today to the original formation of city states and not things like the industrial revolution and the age of AI and stuff like that. And so I'm fascinated and I'll kind of I'll come back to you in a couple of weeks when I'm I'm done because I think that'd be fascinating.

Taylor MacDonald: Huh. Yeah, that that sounds really interesting. Yeah. Please ⁓ please give me an update when you get it. fascinating. ⁓ let's see. I just got done reading The Art of Happiness by the Dalai Lama, and it was great. I loved it. Highly recommend it. Yeah, it was wonderful. He's so smart. Gosh. It's just it's unbelievable how how do you think people get to this stage of enlightenment?

Sean C Davis: Did I no, I haven't done that one. It was good. All right.

Taylor MacDonald: Like what does it take? You know? Is it just endless patience? Is there actual wisdom? Do you just need to eat a bunch of acid and like lose your mind in order to to find a space cognitively where I don't know, it's it's the art of it's so it's really interesting. So the main premise of this book is that ⁓ Human life existence, it's all about suffering, right? And the meaning of life, according to the Dalai Lama, is to find happiness. It's so simple, right? And the best way to find happiness, I haven't figured that part out. Like ⁓ that's in a different book, apparently. But but yeah, I'm working on it. ⁓ sorry, I went down I'd getting a little off track here. So, okay, so tell me a little bit about ⁓ what's going on with the demo. So one thing that just just reminder for our listeners here, we had talked about

Sean C Davis: Yeah.

Taylor MacDonald: How might natural language reduce friction on behalf of users? How might ⁓ we if we were gonna build a content management system ⁓ from scratch, what would that look like today, now that we have this new superpower that is natural language? I also think ⁓ yeah, look let me stop with that and ⁓ You've gone through several iterations. ⁓ maybe could you come back and kind of give us a quick recap of where you stand. I don't know if you want to share screen or if it makes sense or what progress has been made since we last talked, but yeah, where should we take this conversation, Shauna?

Sean C Davis: Mm-hmm. Yes, let me let me share here. And I I feel like my my brain's going in all sorts of different directions because I've been I've just been like hammering away on a bunch of different projects recently. And ⁓ so I'll share some anecdotes as we get into here as well. Also, we're gonna get a little sneak peek of what the website might look like. ⁓ okay.

Taylor MacDonald: Ooh, that's exciting. Yeah, I think we should probably build that get that thing out. It dawned on me we don't really have any feedback loop for listeners, right? Like you just need to know my personal email address, which that's not very helpful. We kinda need a contact form somewhere so we can get the the hate mail.

Sean C Davis: We should, yeah. ⁓ yeah, that's a that is a great point. Yeah, yes, yeah, bring on the hate mail. Okay, so here's well. Let's go. I wonder if I go full screen while it works on it. Yes, there we go. All right. So here's where we were originally. ⁓ I think two episodes ago we ended and you were like, that's interesting. And to be fair, I think that to myself, I think that I was I was using it as an excuse to see. Hey, could I build a desktop application? And the answer was yes, but you don't need to. And ⁓ so essentially what we had here, this is by the way, this is a dramatic simplification just to facilitate the conversation. But essentially what we had done was we had this desk would be that desktop application. And so in theory, you know, you've got your it's running on every individual machine. but I think the ⁓ we're gonna see here is ⁓ bigger ⁓ is the way that I went about designing the whole architecture was really designed toward ⁓ know it's not I I I I I don't want this to come off as negative toward Netlify because I feel like Netlify does an amazing job of this and I did not do a good job of it, but The the stance that we have often at Netlify is you just bring whatever you want and it's gonna work. And the reality is that is that I think that is an awesome approach and it is incredible that we've been able to pull it off there because it takes so much effort to approach a solution in that way. And so I think it was kind of like my brain was working in that way. And so when you're like, ⁓ yeah, you've got this one client that uses Contentful, and you got this other client that doesn't care, and you got this other client on dado. Then we should just support all of those things and and that's where we started from day one.

Taylor MacDonald: Can can I just ⁓ tail like attack on to the end of that? That my wife came to me the other day and she's like, Netlify's the greatest product ever. I should also mention we're not sponsored by Netlify despite the fact that Sean is employed there. I thought it was fascinating. I've been talking about Netlify since twenty seventeen or something. And she the other day she was like, my god, this is amazing. Did you know how amazing Netlify is? And I'm like, ⁓ my goodness. Are you kidding? Of course I d yes, it is amazing. ⁓ ⁓ figured out that you could do kind of voice command and or you know, Claude to deployment. Like she put all those pieces together and now she's de like vibe coding apps and sending them up to Netlify directly from Claude. It's awesome. I'm really proud of her. ⁓

Sean C Davis: Did you know? ⁓ nice, nice.

Taylor MacDonald: Yeah, it's a great it's a honestly what we should do is we should write like a case study or something or you know, like like an interesting perspective from she is not a soccer mom. I'm just using this as a ⁓ term to group people in that stage of life, but like she's very busy, very chaotic schedule, and I think the ability she's also running a business, and so the ability to be like, you know, whip up this report and ship it has been a game changer. And I think that might be an untapped opportunity for your marketing team. Anyway, that's

Sean C Davis: Yes. Love that.

Taylor MacDonald: ⁓ okay, so just to be clear, what we're looking at here, the stuff in red at the top that says desk, it's kinda off screen right there. Yeah, yeah. ⁓ so these six things, these represent individual client machines like laptops or desktops or whatever that are running a piece of software that you're calling desk, but it's really just like a desktop application that was written in Electron, I'm guessing. So you use like TypeScript or JavaScript or something like that. ⁓ is that correct? For the electron piece? Okay.

Sean C Davis: Yes, but the the framework was called it's ⁓ Tari, I think is what it is. Have you heard of this one?

Taylor MacDonald: ⁓ yeah, T A T A U R I, is that right?

Sean C Davis: Yes, that's it. That's it.

Taylor MacDonald: Okay. Yeah yeah, I am familiar with that one. Okay. Now did you choose that or did your agent buddy Hey, have you ever named your agents? Did your agents have names and personas?

Sean C Davis: so yeah. I've I I do that when I have specific ones that I want to ⁓ work together, but when they're individual instances and they're not communicating with each other not as much. ⁓ actually that okay, this could take us this could take us way off topic, but could be really interesting. I built another desktop app and all of these agents do have ⁓

Taylor MacDonald: Yeah, okay.

Sean C Davis: They do have identities. I will let's say if we have time at the end, I can give you a little sneak peek on that thing.

Taylor MacDonald: Okay, all right, I like that. Alright, so okay, so we got all these instances of this desktop application. The orchestrator here is the server engine. This is your API layer, right? Where ⁓ is that correct? Where if I'm doing something in the desktop application. Okay, fair. Like yeah, I noticed you had that. Obviously very scalable. How many clients do you have running on this, Sean? Anyway.

Sean C Davis: Yep. Yes, and and just running on the Mac Mini today.

Taylor MacDonald: That said, the orchestrator is your API layer that is doing the the the massive load here. Like it is ⁓ updating the database, it's spinning out new environments, it's communicating with the LLM to translate your instructions into something, you know, of value coming back. Is that correct? Is that I'm gonna think about this right. So this is like a what is it, an XJS app? Okay. And today it's running on your Mac Mini, but you could just as easily deploy that to Nellify and then you could point all the desktop applications directly at Nellify. Is that accurate?

Sean C Davis: Yes, exactly.

Taylor MacDonald: So The is a server rendered application that does all the transactional with the database, kicks off deployments, that kind of The framework down under client A site and client B site. So these are the front end pieces that are consuming information from the back end, which is largely facilitated by this orchestrator layer. Is that correct?

Sean C Davis: Yes. And and so to go back to your ⁓ right before we we got cut off there was ⁓ it's this it wasn't really even an application locally. It was just it was just services running on the Mac Mini and then used a Cloudflare tunnel to expose specific endpoints. so I mean I guess it would have been an ⁓ but I think it was just like a simple node application, just listening for specific endpoints on one port.

Taylor MacDonald: I see. So I'm a little bit confused with your flow chart here. You've got Netlify and Contentful, for example, attached to client A. And I try not to get too hung up on arrows and flow charts, but it's sort of weird. Like in this ⁓ visualization, it almost looks like client A site is pushing content to Contentful, but that's not really what's happening. Is that a fair statement?

Sean C Davis: No, the error the arrows are wrong. So it'd be it'd be more like all right, client B user one says change this thing, hits this orchestrator. And I was not going as much with the flow and more of just like let's look at the architecture as a whole. But yes, in general, client B user one hits the orchestrator. Orchestrator.

Taylor MacDonald: Okay. Sorry, I don't mean to nitpick I don't need to mi I don't mean I don't mean to nitpick, but the so the orchestrator layer is actually pushing and pulling content from Contentful and then interacting with Netlify to start new builds. Is that accurate? Okay.

Sean C Davis: Exactly, exactly. And then, you know, we went through all these iterations in terms of like do you need instant previews? Are you rebuilding what's happening here? But I the the essence of what I really wanted to do was kind of keep this should this it's probably hard to see there's like a box around each of these stacks. And the idea here is that yes, okay, you have one orchestrator, ⁓

Taylor MacDonald: Mm-hmm. Yeah, I can say.

Sean C Davis: Everyone's bringing their own stack. So you've got to consider all these different frameworks. Like here's a Next site, here's an Astro site, different content sources. Like here's a here's Contentful. Here's this one that we're like, okay, if you if you don't have a CMS, well you can use our database. You can host on Netlify, hosting on Cloudflare. And it was like, okay, just to show one simple example, there are so many moving pieces in this. And that was I think that that where that

Taylor MacDonald: Yeah.

Sean C Davis: You know, where you're like, yeah, this is it's interesting, it's also ⁓ complex for an MVP sort of a thing.

Taylor MacDonald: Correct. Yeah, completely completely agree. You know, and I've been thinking a little bit lately about like the value of this podcast and what are we doing here. And a lot of you know, there's a lot of discussion in the community about college grads or entry level engineers that unfortunately are subject now to this generative AI world where that, you know, no business is going to be comfortable paying well, few businesses I think would be comfortable paying an ex a a significantly sized salary. that could then you know be summarily replaced by a few tokens per month. It's a really unfortunate set of circumstances, right? And so I keep thinking about like what are we doing here? And a lot of what we're doing here is just kind of talking at a very high level, maybe to educate some of those those folks so that they can get a little bit of that higher level thought and perspective and some of the things that you would get in a normal ⁓ office environment. And so when I looked at this, what I wanted to do is spend a little bit of time talking about that. Sure, y it's works. It totally makes sense. I think sometimes as engineers we get ⁓ I don't know, distracted by the shiny object? And I think in this case, maybe Electron was that shiny object. ⁓ that said, it's cool, it works, now you learned something. how have you reflected on it seems like you've reflected on a statement because you put this flow chart together. What do you think the next steps are? Or how might you how might you reduce some of that complexity and or ⁓ I don't know, increase the value of the client investment here? ⁓

Sean C Davis: Mm-hmm.

Taylor MacDonald: 'Cause it's easier to maintain, you know, it's more stable, more testable, those types of things. What are some of those concepts that you would apply now that you have this somewhat sprawling architecture to kind of rein some of that back in?

Sean C Davis: Mm. Yes, yes. All right, let's let's work through that. And you you also you jogged a thought in my mind about the let's come back to the ⁓ the the entry level folks. I've been thinking a lot about that recently and I I wanna pick your brain on an idea I had earlier. But don't gotta gotta avoid too many rabbit holes. This stuff it gets ⁓ this is how you and I communicate, right? Like we we have a goal and then

Taylor MacDonald: Okay. Every time somebody says pick every time somebody says pick pick your brain, I was thinking of that scene in Nightmare Before Christmas where the dude like unclasps his the his cranium right and he like scratches his brain, gets down in one of those cracks. Sorry. Anyway, please continue. Pick pick away.

Sean C Davis: Mm-hmm. Yeah, it's ⁓ I know exactly I know I know exactly what you're talking about 'cause that is a a memorable, memorable visual. and

Taylor MacDonald: Are you gonna see Toy Story Five?

Sean C Davis: Probably. I also I got I I just I'm distracted now 'cause I my my dad is having an operation today and he's gotten into he's gotten into building Legos recently when he's if if he's like, Yeah, it's cold or you know, this or that and ⁓ I've been sharing some of my sets with him but he was out of I think he was out of builds and

Taylor MacDonald: ⁓ my goodness. Yeah.

Sean C Davis: My mom was like, Can you send one up? But we couldn't make the travel work and so I just I ordered and sent him the ⁓ Nightmare Before Christmas scene. There's this big like all there's like four different scenes in this set. It's it looks really cool. I'm excited to see it come together.

Taylor MacDonald: Awesome. Legos. Legos are so freaking expensive. I should have bought l stock in Legos back in like the eighties and be a bajillionaire now.

Sean C Davis: ⁓ yeah. Yes. That's why what we do is I ⁓ I I bought a bunch during the pandem I spent all my pandemic money on Legos instead of a house. And ⁓ and sourdough starter. Yes. So now I and I'm I'm really particular about the the Legos and I want the experience of building the second time and third time and fourth time to be as close as possible to the first time. And so

Taylor MacDonald: ⁓ I thought you spent it on sourdough starter supplies and equipment. Me too.

Sean C Davis: I put a I build it, let it sit for a couple months, kids play with it. And then I am like super meticulous about taking it apart and putting it into bags that I number so that you can do it the same way the second time. And so this is this sounds so nerdy. But the benefit of this is that then I can take the box and give it to somebody like my dad. And he doesn't have to spend $300, but he can get the you know, a close to a ⁓

Taylor MacDonald: You nerd.

Sean C Davis: three hundred dollar experience, so to speak.

Taylor MacDonald: I love it. That's a great idea. That's so awesome. What a what a great way to repurpose to the most memories. ⁓ okay, I have many thoughts. One, and I know we're off topic, but whatever, this is fun. The primitives that you guys talked about. ⁓ these platforms, Nelfine, you guys are always talking about your primitives. Those are just Lego blocks, right? I mean, in one way to think about it. that concept appeals to me, right? Because I I think about you know, I mean, even UI components, they're just Lego blocks. You're just snapping them all together.

Sean C Davis: Mm-hmm.

Taylor MacDonald: Make something. It's kind of an interesting way to simplify some of this. You know, I would imagine that if you're an entry-level engineer and you come in, you listen to this podcast, you're looking at this flowchart, or you're like, I don't even know where to get started. I mean, at the end of the day, you've got a whole bunch of Lego blocks. You're playing a game of Tetris. Every project is a new game, and you're trying to figure out how to fit the little blocks together, whether on a user interface or from an infrastructure standpoint. That that simplicity appeals to me. Okay, the other thought that I had is have you ever heard of Nintendo Labo? I think Labo Labo, Labo? I don't know how you pronounce it. You ever heard of this? L A B O. I don't even know if it's still a thing, but this is hands down one of the most creative and fun, cool activities I have ever done with my children. So we have a Nintendo Switch, which I'm assuming you have. If you don't, you should. You're Okay. So Labo

Sean C Davis: Mm. How do you sp how do you spell it? No. I do have one. Yeah.

Taylor MacDonald: is basically a box of different cardboard, like pre die-cut cardboard, perforated cardboard, ⁓ of sheets of cardboard, I guess. ⁓ you pop all the little pieces out, ⁓ it comes with a game cartridge. You plug the game cartridge into the switch and then the game itself, or at least the experience when you first open this game on your Switch, ⁓ it will walk you through the steps to fold and assemble these cardboard ⁓ ⁓ something ⁓ ⁓ ⁓ my god. It is so cool. So one of the best activities that we did was we built a frickin' piano out of this cardboard stuff. And what it's it's awesome. It's so great. Like whoever engineered this is the most brilliant human, which is not surprising coming from Netlify or so from Nintendo. Sorry, N word.

Sean C Davis: This is awesome.

Taylor MacDonald: Y the piano, it's so great. What you do is you take one of the little Joy-Con remotes, right? Which are the two little things that you separate from the machine when you're playing like tennis, you know, you've got the Joy-Con in your hand. So you take that little thing and you they build these little like boxes that you can slide the Joy-Con in. And I don't know if you know this, but on the end of the Joy-Con there's a camera. ⁓ and I have never I mean, I'm not a I'm not a huge video game dude, but ⁓ you know, my son is, and he's playing

Sean C Davis: Mm-hmm. I didn't know that.

Taylor MacDonald: Less of the touchy feely, super cool, innovative ninth like Nintendo branded stuff and more of the other, you know, older, more mature games, that kind of thing. anyway, I have never played a game on the Switch that actually took advantage of this camera. But if you take those Joy-Cons and you like flip, I think it might be on the bottom, there's like a little black kind of vented sort of window, and it can see out of there. And so that technology is used in the Labo. do is you basically you know you plug your joy-con into the constructed piece and on the inside of these like in the case of the n the piano there were like you know five or six keys like individual cardboard keys you know long ones right and I think they also had the minor keys and all that stuff too. Well inside the piano ⁓ base or whatever at the end of the key you wrap a little piece of reflective tape around it.

Sean C Davis: Mm-hmm.

Taylor MacDonald: And somehow that reflective tape can get picked up by the camera in the Joy-Con. So when you put the Joy-Con in, it's watching the keys move on the inside of the piano. So you hit the key, you know, you hit it outside and it like toggles up, right? But the reflective tape on the end of the key is now like being captured by the Joy-Con camera and it can react. And because it's all wired up to the switch, once you turn it on. You can literally sit there and play piano through your TV using a bunch of cardboard. Mind blown. It was it's so cool. Or if your dad has a switch, order this for him. He will lose his I know your dad, he will lose his mind. It is the coolest thing ever. And they've got like ten different sets or something. I mean there's

Sean C Davis: That's genius. It's genius. Yeah, he does have one. That's a great idea.

Taylor MacDonald: Yeah. Now that's Father's Day. Just be like, ⁓ sorry man, it was delivered late Send it to him tomorrow, like Amazon this thing, overnight tomorrow and be like, It finally showed up and then you're like the greatest son ever and he has plenty of things to occupy himself while he's recuperating.

Sean C Davis: Ha ha ha. That's that's awesome. We got ⁓ there's a there's a game for the iPad. Well you have several of them, but they're called Osmo O S ⁓ O. Have you heard of these?

Taylor MacDonald: Mm we've got one. Or and we've got a whole bunch of my kids have outgrown it. But ⁓ yeah, it we it's like little shapes. So what is it? It's like you take your iPad and you stand it upright vertically. And because it's got the camera right in the middle of that thing, you can put this little it's brilliant, man. You put this little plastic contraption over the cameras, like over the eye hole, and it's got like an inverted mirror that causes the camera to have a view of your desktop in front of you.

Sean C Davis: Yes, exactly.

Taylor MacDonald: And so then what you can do is you can take a piece of paper and you can draw on the paper, or there's other like, don't they have like little there's like little puzzle pieces you can put together to play different interactive games? But what you're doing on the desktop is then reflected on the screen in front of you, and it's all because of this mirror, it's just able to look directly down. Man, people are so freaking smart. Who comes up with this? This is just brilliant. I'm so in awe of that. Anyway.

Sean C Davis: Mm. ⁓ yeah, yeah. I agree. I agree. Okay, so back to this. So yes, to answer your question. All right. So there's a phrase that I have liked to use for years that I apparently just completely ignored here, which is I I like to tell people and try to tell myself that we should opt into complexity. So we should start with simplicity.

Taylor MacDonald: All right, we digress.

Sean C Davis: And add the complexity only as we need it. And in that version 0.2 that we just saw, that was way more like complexity, just we're we're bringing it on ourselves as a byproduct of supporting more. And so the shift here, after we talked a couple weeks ago, was way more about let's let's more strictly define what the happy path is for this MVP. And then Let's let users opt into that complexity. So we'll build those escape hatches, but we're going to say if you want a great experience, just do this thing. And so I'm I'm I would say I'm I'm about I finished the first whoops, finished the first rough run yesterday. So I think we're we I would consider it about two-thirds of the way done ⁓ before we can actually wire this thing up and use it. But the concept here is that every that you can just pick things off the shelf and because we have AI, it can it we can we can build this whole system that can be exactly what a client wants. So the idea here is that there's this hub and all the hub is is just ⁓ it's an application with an API and a database behind it. It's something we can we could Pro we'll host it on Netlify, but you know, we could deploy it anywhere. But it's something that everybody can use. So it's our kind of entry point for anybody who's working with the system. And it is essentially this orchestrator. And so then somebody wants to create a new application, then they'll come and I the vision I have, which I think will take a little bit to get there, but you and I have talked a little bit about like.

Taylor MacDonald: Yeah.

Sean C Davis: Streamlining that discovery process. So the idea is that there would be this base level template that is already built to work. There's a there's a base set of components. Anybody can pick these up and use them, or you go to this hub and you're like, I'm ready to start this new website. And then we can work with AI to say, okay, where's your content today? What is your brand system? Yada yada yada. And then we'll spin you up an MVP. But we're but we're going to spin you up that MVP in this really narrow set of criteria so that you're using the SDKs that we have given you and the A and the APIs that we have provided. So all of these sites essentially work the same, again, unless you want to escape from that for whatever reason. and I I'm not showing it here because I'm going to I want to iterate on some of the infrastructure, but the concept is There's there's basically like I think three no, let's let's call it two. There's I I think there's two types of agent interactions. And I'm starting to split those out here, which might seem complex. I eventually want to bring them together and I'll explain why. So the the first type of of interaction is the really simple one where These open source models or Eastern models can be they're they're gonna be more than good enough. And that is I need to change the heading on my home page, right? We don't need Opus to do that. We can use Quen or Deep Seek or whatever. And so this hub is gonna route that request and make use of open router and one of those really cheap models to just resolve what the human is trying to say and how to turn that into a deterministic request. But then

Taylor MacDonald: Love that.

Sean C Davis: There's way deeper work that has to happen. Like I want you to add a new component or I need you to migrate all of the content from my old site to my new site. And so today that's gonna sit on the Mac Mini because we can take advantage of either my Claude or Codex subscription, which I've I've been ⁓ using using heavily. And I eventually though, I'd love to get that onto some it's some sort of Dockerized container or or something like that, but the still the hub is the entry point. And then we can and then we can get more tight with those opinions too and basically be like, all right, how do we build in a review workflow, etc.? But one of the things, last thing I'll say here is because we've then provided these tools, like here's Astro off the shelf, here's this ⁓ component library, here's this SDK of how you can interact

Taylor MacDonald: Yeah.

Sean C Davis: it it's a it creates a surface for your website so that your users can update things that we can also then already have it pre-baked to support these preview environments so that when you say update this thing, we can go boom, instantly you can preview it, but it's not available in production yet. And so I it's ⁓ I don't know. I'm just really I'm I'm really glad you and I sat down and worked through it a couple weeks ago because I feel like that gave me a little bit of freedom and focus to just be like, just build the one flow and make it great. And I feel like we're on our our way by like shifting that model.

Taylor MacDonald: Okay. I have so many things I wrote down like a trillion notes here. let me let me think about the so one of the most important things that you said is has to do with this happy path, the concept of the happy path. You hear about this all the time, right? Like

Sean C Davis: He

Taylor MacDonald: The happy path, when we talk about that in terms of like digital marketing or lead conversion, it's usually like what are the steps that the user has to follow to get them to do what you want them to do. And your particular case, happy path, I would think, probably has more relevancy when you consider that in terms of developer experience. The happy path for the developer is I click a button, I have something, and now I'm looking at it, and I can I can actually start doing something useful with this. And then tomorrow, if I wanna grow from there, I want to expand, I wanna fork it and do something else, I I still have that power. I'm not limited. Right? This is a little bit different than ⁓ I don't know. When you start like traditional I remember when I was in early in my career and I wanted to learn how to program I I started my career in the late nineties. I mean I was tinkering around with websites, but I didn't understand it and I wasn't doing it for business, so it was fine. And I thought I was gonna be a designer. And then Somewhere along the line I realized that ⁓ you know, developers kept screwing up my designs. They'd like to send back stuff with broken images and I got annoyed. So I was like, screw this. I'm just gonna learn how to build this stuff. And it turns out it's not that complicated. But anyway, the happy path is an important concept and the reason I'm kind of zeroing in on that is because it reminds me a lot of the methodology that ⁓

Sean C Davis: Yeah.

Taylor MacDonald: That DHH ⁓ had at 37 Signals years ago when he built Rails. It was this highly opinionated software. You know, back when I first started doing development in the late 90s, it was the Wild West, man. I mean, if it worked, dude, you're like, hell yeah, like we're going to lunch. This is awesome. I actually wrote JavaScript and it worked. There weren't concerns about security or consoles. I mean it we were just like really flying by the seat of our pants and then over time different systems and services propped up and I remember when I first started writing action script like you know my mind was expanding and thinking about all the stuff you could do and and yet it was still like batteries not included kind of world. You know, every time you would sit down, you'd open up Dreamweaver or whatever it is and you're looking at a blank page and you're like, okay, I have to figure out how to put together all the Lego blocks. But at the time we didn't have Lego blocks. We just had a whole bunch of like like shitty ⁓ man, I was trying not to cook.

Sean C Davis: Mm-hmm.

Taylor MacDonald: Cuss. Dang it, I'm screwing this up. We had a whole bunch of I'll go back and bleep that up. We we had a whole bunch of like I remember I bought this JavaScript book in like ninety-eight. And it if you go back and read it in 2026, it's so wrong. Like everything in there was rife with error and mistake and of course it was a sign of the times too, so it was like we're talking about Mozilla and you know Alta Vista and all this stuff. Anyway, the reason I'm

Sean C Davis: Got it.

Taylor MacDonald: belaboring this is because there is a real value for the engineer, specifically for the more junior engineer, to select a framework or select a solution that comes with batteries that has an opinion. The thing about Rails, it took me a long time to figure out, it's like if you work with Rails and you're doing it the Rails way, the opinionated way, it is a frickin' breeze. It is so pleasant. It is like a wonderful experience. I remember

Sean C Davis: Mm-hmm. Yeah.

Taylor MacDonald: You know, not only like the syntactical sugar of Ruby as a language made it expressive and easy to write and fun to format your files and all that stuff, but there was so much power. You just plug in this gym, you do this thing, write this boilerplate code, stuff that in your model. Here's a little update for the controller, a boom, it just works. And I thought that was so powerful. does that resonate with you? You talked about opting into complexity. That is exactly what that it's like, hey, I just want something that works. And I I fear that the more junior engineers that are getting into the game that may not be able to distill the differences of this library and that dependency and this requirement, it can feel so overwhelming. That I think there's a ton of value in the way that you've kind of approached this, you know. Granted, ⁓ you're you're you're speaking more to the client and I'm speaking more to the engineer, but it's the same concept, right? It's like I don't want to get started and I don't want

Sean C Davis: Yeah, yeah, yeah.

Taylor MacDonald: to I don't want a ton of friction in between me and the outcome. I need to see value early or else I get discouraged and quit working.

Sean C Davis: Yes, yes, exactly. Like we want that magic to come right away and and then be able to learn and kind of you know, experience this just in time learning. So I need to do this thing and I wanna make sure there's a path to do that. And I actually this is I put a video together on this topic, ⁓ I think it's actually gonna come out today. And it's it's going through this this story of my experience. So I came across Rails really early on in my development career. And yeah, I it it was fast to learn. And within a couple of years, I would say, I was building apps so fast because I because Rails has its opinions. And then I built a thin layer ⁓ of my own libraries on top of Rails. And it was like I could just I could just fly. And then when we get to this decoupled world Wiring up these services just took so much time that I went from building my own tools to basically buying all of these tools. But now I'm back to building the tools, partially because of AI. But I think this is a great opportunity to be like, yep, follow this path. This is a layer that sits on top of infrastructure that is ultimately very flexible. So take it as far as you want. You can get going. And then when you want to dig in and you want to learn more.

Taylor MacDonald: Mm-hmm.

Sean C Davis: You can do that. And that okay.

Taylor MacDonald: Do you fear do you fear the loss of rest do you fear the sorry to interrupt, do you fear the loss of restraint? Like just because you can, does that mean you should? You know, back in the day we used to spend a ton of time like belaboring, like, well, if I add that dependency, it's gonna do this, this, and the other. And I feel like nowadays engineers are like, YOLO, who gives a crap? Like, let's just go. Put it in there, you know, make it work.

Sean C Davis: Yeah, yeah. I think I'm exploring f that line. I'm I'm not I'm not totally sure where it is yet. I but I think that yes, I think there's gotta be some level of restraint 'cause you can't you can't manage everything. ⁓

Taylor MacDonald: Yeah, but the altern like okay. It's so interesting because for many years my experience I used to be terrified and I I just want to come back to this statement. I wrote down lean on your cohort. Like you and I talking through some of these challenges has probably shifted the way you think about it. Not because I have all the answers or anything, but because just talking. Just th there's a whole concept in programming called like ⁓ duck typing or whatever, where you pull a rubber duck out.

Sean C Davis: Yeah, yeah.

Taylor MacDonald: And you set it in front of you and you explain the problem you're trying to solve or how you solved it to the duck. Which sounds ridiculous, right? It's not though. The exercise forces you

Sean C Davis: Mm-hmm, mm-hmm.

Taylor MacDonald: to think about it from a different vantage point. I think when we get down in our code, we're down in the weeds, sometimes we're missing the force for the trees just because we're so hyper focused on the outcome. Just make it work, make it work, make it work, make it work, and after like six hours of that, ⁓ it can be very discouraging, right? Or or y you know, you finally let's say you finally get it to work and you're encouraged and then you like load up the browser, refresh the price, something goes haywire. It's like, right. There's an opportunity to slow down a little bit. And I guess the The recommendation I would make to more junior local engineers is go find the networking group. Go find your Sean. Go find your buddy and talk to them about the problem. And in the absent of that, tell your dog or tell the duck on your desk ⁓ what the problem is. I think there's a lot of value in that. Anyway, I interrupted you. Go ahead.

Sean C Davis: Yes. Yeah, yeah, a hundred percent. And I think, you know, on that note that ⁓ I think it's it's always more helpful to talk to a human for sure, but I have in in the absence of having a human available to me, I've also found that telling telling an agent to be a to play devil's advocate also really helps me think through so it's not a duck. It's I mean we could we could call it a duck if we needed to. but

Taylor MacDonald: Did I tell you about the entropy thing? I had a conversation with a dude. I had a conversation with a dude a couple weeks ago and he was telling me how they do like ⁓ I think he told me it was like focus groups kinda. Like they'll they'll have a whole agent swarm.

Sean C Davis: I ⁓ n no. I don't maybe?

Taylor MacDonald: spin up ⁓ an example of a focus group to like test and validate ideas or something. But the problem was I guess that the focus groups were always returning fairly generic and fairly s similar responses. And they figured out that if they injected something completely random into the prompt, you know, just like tangentially related but complete, you know, like off the wall

Sean C Davis: Okay.

Taylor MacDonald: that that would incur you know that and they for every agent that was participating, you know, in the focus group, like if they injected something different into each agent's like system instructions, that that would cause them to come back with significantly more creative output. It's it's effectively like engineering entropy. ⁓ I forget the way he described it and I need to like sit and sort of reflect on the conversation and think about it. But I the sense that I got was like you could just inject

Sean C Davis: Like what? What would be an example?

Taylor MacDonald: some sort of like I don't know you know how sometimes you give agents personas, you know, like you are so-and-so from here and here, this, that, and the other. That like that exercise of of adding more rich and uniqueness to the prompt results in more creative output because it sort of ⁓ it's injecting randomness Right, entrop computers run on entropy, right? Have you spent much time thinking about this ⁓ about this concept? So entropy is the I'm gonna bastardize this, but entropy is like the computer equivalent or the ⁓ the process of generating randomness or seeding random randomness into calculations.

Sean C Davis: Yeah, yeah.

Taylor MacDonald: ⁓ and it is important because otherwise computers let me s hold on. I should probably Google this just so I don't screw this up. Here we go. A definition of entropy. It's a measure of disorder, randomness, or the unavailability of energy to do useful work within a closed system. In a natural world, systems naturally evolve from states of order to disorder. This is random like so if you think about like organic nature, life. the waves in the ocean, it is inherently random. There is nothing that will ever govern the position or movement or fluidity of those waves, except the wind and all these other factors that are so unique that they can't be accounted for. There used to be a blog post by ⁓ Cloudflare. I don't know if they still have it up, but they put a blog post about how they generate rand like so they need to generate randomness at scale.

Sean C Davis: Mm-hmm.

Taylor MacDonald: So that they can do like cert you know, like encryption stuff and all that kind of things. And the way that they did it's brilliant. They put together a wall. Do you remember the story? They put together a wall of like two hundred lava lamps. And it's awesome. The lava lamp thing, yeah. And so what they did was they took a picture of the wall. So you two hundred friggin' lava lamps. A lava lamp is the most organic thing you could ever think. I mean, it's brilliant concept, right? Because there is absolutely nothing that governs the fluidity inside a lava lamp except

Sean C Davis: ⁓ yeah, yeah, yeah. The lava lamps, yes.

Taylor MacDonald: Heat, you know, and whatever forces like live inside that thing, meaning like you can look at that at any moment in time, and there's a bubble here, there's a bubble there, there's this big thing is coming up, it's separated, and that looks cool, or whatever. That's the most random thing they could come up with. A lava lamp is just organic and random, and so then they put 200 of them together on a wall or something and took a picture of it every 10 minutes. And every 10 minutes, that picture was then distilled down to some sort of random value. And I don't I forget how they did it. They go into great detail in the blog post talking about how they do it. But anyway, entropy is real important. let me see. ⁓ computer science and telecommunication, ⁓ measures the amount of uncertainty, unpredictability, or surprise that is inherent in a data message. ⁓ highly predictable message would have low entropy, where an entirely random data, like static, for example, would have high entropy. ⁓ so I think the idea is if you can introduce random into the thing, that helps encourage the non-deterministic creativity coming out on the other end simply because you seeded it with something unique.

Sean C Davis: ⁓ Okay.

Taylor MacDonald: Pretty cool com pretty cool concept. sorry I took us off track again. Hey, can I ask you one question though? You mentioned earlier Open Router. Could you describe what OpenRouter is and why you are using that and how that may or may not be preferable, particularly for junior engineers, because junior engineers don't have money.

Sean C Davis: Yes.

Taylor MacDonald: You know, one of the things that I've noticed is like it's easy for you and me to throw ⁓ like you mentioned that you're maxing out your Claude and Gemini subscriptions. How much do you pay per month for those two subscriptions? Are they personal subscriptions or does Netlify foot the bill for that? ⁓ wait, don't answer that. If don't answer that if it pisses off your CISO. Sorry. Whoops. anyway.

Sean C Davis: ⁓ God. Yeah. I mean I No, I ch ⁓ they they get charged to Netlify. No, no, they're they're ⁓ we and and and we're working on, you know, it's it's it's still the Wild West. ⁓ we do have a formal contract with So company account there. everything else is experimental, but I like playing them off of each other. So I've got the hundred dollar a month codex one.

Taylor MacDonald: Okay. Well let's talk about Yeah. Yeah, yeah. Hundred dollar a month. Okay.

Sean C Davis: And then then the and then the max on ⁓ on Claude. But you know it's they're so subsidized.

Taylor MacDonald: Is it the five so I've got the yep. I think our delay is back. I think our delay is back. Anyway. Yeah, no, sorry, just to be clear. So you've got basically twenty four hundred dollars worth of agents that you're running. Right? Of a year. Like if if you had a a assuming you were paying for yourself, right? Like you're a startup or you're a solopreneur, you're a contractor or whatever, you've got two different service providers, one in anthropic, one in open AI.

Sean C Davis: Sorry, go ahead, go ahead. Yeah. Yeah, right.

Taylor MacDonald: Hundred bucks enough for each twenty four hundred dollars a year, that is insurmountable. I mean for a lot of folks, like that is a lot of cash. So why what is open router and how could it necessarily alleviate that cash crunch?

Sean C Davis: Yeah, so the it's now I'm doing I'm doing a lot of heavy lifting and and testing for part of the job. So part of the reason I have these subscriptions is because I my I use them for the job. ⁓ and I think when it comes to writing code, you know, having these these ⁓ what do you call these harnesses like clawed code makes it really much easier and faster. But you know that that hundred or two hundred dollar a month clawed subscription in the equivalent API cost, because I'm mostly using Opus, I've just looked at it, the last 30 days it I would have charged almost $1,900 just in 30 days, which is outrageous. Outrageous. ⁓ it's a a lot of money. All right. So some of these other open source and eastern models

Taylor MacDonald: What? That is outrageous.

Sean C Davis: are fractions of a the the cost of that. And it I haven't explored the coding part of it, but when I'm just like typing and messaging, the they're they're so much cheaper. Now the context windows are small, but that's okay. They can be smaller because they don't need as much context when you're not doing a bunch of heavy lifting. And so open router is a great tool because you have this one account

Taylor MacDonald: Mm-mm.

Sean C Davis: You have one API key, or you know, many, doesn't matter, but essentially you can have a single API key and you connect to open router and you can swap in and swap out any model. Now the caveat there is you're paying the API costs. So for me, as somebody who has a Claude and Codex subscription, it does not make sense for me to use any. anthropic models through that because they're going to be really expensive because I can just use them through the subscription, which is getting subsidized. But when I need to do things either on the fly or in the cloud or whatever it may be, I can use these much, much cheaper models. And I am doing I'm like I'm pretty constantly chatting with with ⁓ with tinkering with this app that I've I've been messing with, ⁓ which I'll I can show next time. Using exclusively open router for that, I over the course of three weeks I've spent like two dollars or something like that. It's crazy.

Taylor MacDonald: Yeah. That is crazy. And I it's so interesting. I think it's not surprising, you know, marketing is a really powerful tool, but I think most American consumers of this technology are not aware how many other opportunities or different solution providers exist out there. An open router is a great way to explore some of that stuff. And they have really

Sean C Davis: Yeah.

Taylor MacDonald: Really interesting ⁓ data on the website, which you can if you go to open router.ai, I'm gonna share my screen too, because like some never mind, I'm not, because I don't have haven't given it permission yet. Anyway. ⁓ you go to ⁓ slash models, you can kind of like look through, and they are filterable by the type of work you want to do, whether you're working with video, audio, text, images, files, you name it. And then for each one of the models that exist. inside this, you know, service provider, it'll give you context window, it'll tell you how big that is, which is important, right? Because once your context window gets full, then you either have to go through compaction or you just lose data or it starts hallucinating stuff more frequently. It'll also show you the total cost per million tokens. Here's an example. I'm looking at my screen right now. ⁓ Anthropic. Wait, is Fable back? We talked about this last time. The last episode, we kind of screwed up our schedule, and so we actually did not record one last week. When we recorded two weeks ago, we were both like, hey, have you tried Claude Fable or whatever? And we're like, ⁓ I I had not tried it at the time. I think you had. But then about 24 hours later, the government shut that thing down and removed access for everybody. You remember that? We haven't had a chance to talk about that yet. Anyway. Anyway, according to open router, Claude Fable Five.

Sean C Davis: Is it? Mm.

Taylor MacDonald: Is ten dollars for a million input tokens. Okay? Alright, that seems reasonable, whatever. The probably closest ⁓ competing open source or at least a very popular open source model is ZAI GLM 5.2, which just recently released. Zero dollars for a million tokens. Right? You're like, wait a minute, that doesn't make sense. It it's w I mean it's not straight up free, but it's so cheap compared that it's not even it doesn't even register on the dollar scale for a million tokens. It is approximately forty-one cents for a million tokens on input. And then on output they charge you approximately four dollar and forty cents for a million tokens on on the output. And let me go back to f Fable. We'll get the numbers on that one too. I just think this is interesting. I'll tell you what, I'm gonna look at Opus four eight is a little bit cheaper. And

Sean C Davis: Mm.

Taylor MacDonald: I don't even know if Fable is actually I don't know if you can use that or not. I haven't messed with it yet. Fable is five bucks for Okay, for a for input, meaning the words that you're writing that are getting sent to the N the L L

Sean C Davis: It looks like it's still not available.

Taylor MacDonald: For a million tokens, it is five bucks for Anthropics OPIS ⁓ 4.8. Okay. That is five dollars for a million if you use ZAI, which is arguably not as powerful, but pretty dang ⁓ pretty dang comparable in terms of actual you know functions and and output. I mean, it is quite comparable. It is a positive experience that I've had with it for what it's worth, both in coding as well as just general chit-chat. the difference is five dollars to forty one cents ⁓ the same ⁓ of tokens. ⁓ is the that is the scale and difference between these things. What is that? Like ⁓ twenty times difference? Ten times difference? It's it's bananas. ⁓

Sean C Davis: Mm-hmm. I trying to find something where ⁓ I forg I forget what site it was, but it was like there were some cases where ⁓ that GLM five two was outperforming ⁓ and for a fraction of the cost.

Taylor MacDonald: Yeah, yeah. That's the that's what I think is so interesting. You know, like, okay, so Fable gets yanked off the shelves because of a perceived security threat. I won't go down this the conspiracy rabbit hole here on this one. It could be that. It could be other reasons, say political reasons, whatever. it gets yanked off the shelf. I think that it is not a good look for US service providers because these other models exist and I there's probably a discussion to be had around data security and privacy. I just don't I just don't see how I don't see how the technology doesn't catch up. And it gets homogenized to the point where it's like, I don't care. Like I you might be able to, you know, build a nuclear reactor in my backyard with this technology, but I don't really give a shit. I'm not building nuclear reactors. I am building websites. I need I need something that is cost effective. I mean, and you talk about the just constant when we talked about this last time, the constant you know, merger acquisition landscape of all of these service providers that we use, ⁓ then you're forced in To you know, jacked up prices, or you're locked into a contract with a company you didn't even dude. I got a bill the other day from an insurance company, Liberty Mutual. I don't do business with Liberty Mutual, I don't have an account with Liberty Mutual, but I got a $70 bill. So I called him and he was like, he was like, Well, what's the problem? And I'm like, can you show me my policy? Like, I don't understand where this bill came from. It was a merger.

Sean C Davis: Ha what?

Taylor MacDonald: The other underwriting insurance company no longer services my insurance provider and so they just s switched it. They probably sent me an email or something, but I don't read I get so much email is just like a cesspool of of mostly not good information. So I probably missed it. I'm sure legally they notified me. I mean I'm sure they have all sorts of obligations to do that. But it was just surprising to me. You know what I mean? Like we could have included a cover letter and said, Hey, you know, you got gobbled up. Anyway, the point being is that if you opt in to an enterprise level anthropic account, it's gonna be a great experience until the other shoe drops And the investors call that cash back in. I mean it's exactly no offense, but doesn't Netlify do this with their dang Percy pricing model and then they backed off of it and Vercell did the same thing. I mean everybody does this. It's like they get you hooked on the service and then they flip that switch and I don't even blame ⁓ I I totally appreciate that. I just it bothers me that like you can't it's like a gamble. It's like, ⁓ should I do business with this company or this company? Who's most likely to sell out and stick me with the bill, you know, when it's convenient for you know, leadership to do so. And I don't even blame them. It's just as a you know, a service-based industry, it's almost impossible to negotiate that complexity. And open router gives you a solution. Open router solves this because you can use any of the models. You know? You don't like Fable, cost went through the roof. Pivot over to something else. You spend all day playing with different you know, different models and come up with the one that works best for your workflow.

Sean C Davis: Mm-hmm. Yeah, yeah. Exactly. Exactly. And you're paying for what you use. I think that's ⁓ and that's always the case. Yeah.

Taylor MacDonald: Agreed. And man, we're over time, dude. We've been talking too long today. Good chat, Sean. I think it's the mustache that made it so good.

Sean C Davis: Yeah. I I I can't argue with that.

Taylor MacDonald: Hey, I got a question for you. Can I ask a mustache question?

Sean C Davis: Yeah, yeah, yeah.

Taylor MacDonald: Have you ever Okay, so on your birthday every year, you usually dress up ⁓ wait a minute, let me back up. When is your birthday, Sean? Sean has a birthday every four years. February twenty ninth. February twenty ninth, the weirdest how old you now? Counting actual physical birthdays. We seven? Are you eight yet? You're nine?

Sean C Davis: Every four years. It is it is leap day, yes. Only every four years. Yep. Nine and a half. I'm getting old. Nine and a half. Yeah.

Taylor MacDonald: ⁓ my good dude, when's your tenth birthday? ⁓ that's when you turn forty. ⁓ my god, that's a big one. That'll be fun.

Sean C Davis: Yes. My daughter turns ten the same year also.

Taylor MacDonald: That's right, that's amazing. I lo are you guys gonna have like a like a dual ten year old birthday party?

Sean C Davis: Yep, joint party. We're making it happen. Yep.

Taylor MacDonald: That is adorable. Okay, anyway. ⁓ man, I forget where I was going with that. ⁓ the mustache. So every year for your birthday, every four years for your birthday, you like trick out your mustache, and I've seen you show up like fully greased and like, you know, dare I say erect ⁓ handlebars coming off that face. what do you use to grease up your mustache?

Sean C Davis: Yeah. mustache wax, of course. ⁓ it's ⁓ wax is a product. ⁓ probably been a product for ⁓ few hundred years, 'cause I think folks, ⁓ you know, when it when it was actually acceptable.

Taylor MacDonald: What is that like mm ⁓ that's a thing? That's a product? Mustache wax? Is it when you put that crap in your face, how long does it take to wash it out? Like I mean, is it like my dog got skunked the other day and I could not I mean that's like frickin' petroleum. It was like napalm, man, you cannot get that stuff off. Is this like napalm? Am I gonna is am I gonna have like sandy, nasty, waxy crap in my face for like the next two weeks because the product was developed in the seventeenth century?

Sean C Davis: Ooh. I you know, to be fair, I don't know that I can accurately answer that because I'm not allowed to keep it for more than a day. So the way I get it off is usually by cutting it off my face.

Taylor MacDonald: Hilarious. ⁓ really? You have to cut see all right, I'm an old punk rocker and these dudes used to build they used to make mohawks with Elmer's glue. It was the punkest thing you could do. It was insane. It was awesome. It was so cool. It was like, bro, you are committed to the craft here. I appreciate that. But yeah, with Elmer's glue, I mean, that ain't coming out, bro. Like you were shaving your head when you when you get done. I don't want to do that with my face. There has to be a

Sean C Davis: ⁓ my god. Yeah. Holy smokes. Yeah. Yeah.

Taylor MacDonald: a ⁓ a modern version that I'm not like because I've thought okay I have this idea. This is crazy. But okay, so I have long handlebars too, right? And I was thinking like, what if I grew up I also have like really long ⁓ eyebrow or like ⁓ eyebrows. Like they get really hairy and long and they start drifting. And every once in a I'll show up and look in the mirror and there's like one rogue hair that's like three inches somehow that grew overnight and is poking out over my eye. But what if

Sean C Davis: No no.

Taylor MacDonald: I braided my mustache handlebar into my eyebrows. And just created like a whole new thing. No good?

Sean C Davis: I mean I'm not telling you not to. I think that it functionally could get in the way of your your vision potentially.

Taylor MacDonald: Yeah. Well, I'm getting old. I'm okay with that. I I can't even read the Lachlan handed me something this morning and she was like, What is this thing? I'm like, I have to find my glasses. I can't. It's a I Anyway, I'm I arrived. All right, dude. Hey, thank you so much. There's a great chat today. I really appreciate ya. Next week, let's come back and maybe I'll allow you to actually get to the finish of your demo.

Sean C Davis: Yeah. I mean Okay. Yeah. There we go. We're there. We're there.

Taylor MacDonald: ⁓ I had one other question for you before we go. I'll save this for next time. It's a question about Nell Fi CMS. We'll talk about it next time.

Sean C Davis: Hey, we'll see. Yeah. Yeah. I've got ⁓ I've got another question for you too that I wrote down that we'll so we're ready to go, already for the next one. Great.

Taylor MacDonald: Okay. All right, we got an agenda. Okay. Hey, thanks man. I will ⁓ we will see you soon and thanks everybody for listening. Bye.