Rethinking Developer Life and Productivity with Rapid AI Advancements

Zach Stepek:
Welcome to another episode of Expanding the Stack. I’m Zach Stepek here as always with my illustrious host, Carl Alexander. It’s been a very interesting time since press conf. You did an episode last month by yourself because I was dealing with severe allergy issues and a headache that wouldn’t go away. That episode turned out awesome. But this month we are going to be talking with Alex Standiford about Friend of the pod. Yeah, definitely a friend of the podcast and you know, somebody that I’ve been fortunate to call a friend and I know, Carl, you have as well.

Carl Alexander:
Yeah.

Zach Stepek:
Now, Alex, you have built Siren Affiliates, which is your plugin for affiliate. Do you want to talk a little bit about that right at the top so we can get into all the meat?

Alex Standiford:
Sure. So one of the. So Siren is built, I, I think of it as infrastructure, especially with AI and where it’s been going lately and stuff like that. I think that a lot of the tools of the future are going to be things that aren’t just I fix a thing, I solve this specific problem like they used to be. I think we’re running into scenarios now where people are at all different skill levels starting to be more willing and able to try to extend things and they’re going to want to create them exactly how they want to. So I’ve started looking at Siren as infrastructure as opposed to a plugin that solves a specific problem. And it’s always been built that way, you know, to be able to do any kind of incentive program that you want. But the key thing there is that means that Siren has to not just be a WordPress plugin, it has to be able to be its own hosted service that, you know, my clients take, take me on for. Right. So, so I have high touch clients that’ll hire me basically to consult with them, to help them create their sales programs internally at their companies or if they’re running some, some bigger affiliate programs. Right. So I’ve talked to some hosting companies in the past about doing things along those lines. But it’s, you know, any kind of program you can imagine where there’s any kind of structure. I was literally just working on this, abstraction for a set of walkers that makes it possible to walk up and down a different trees for, you know, downline, upline, things like that. So any, any kind of program at all, way outside of just affiliate programs. If it’s incentivizing, if it’s tracked, Siren is positioned to be able to do that. So that’s been the biggest, the big shift there. And as for the tech stack and stuff like that, we can, we could talk about that a little bit later, but it’s, there’s, there’s been some pretty interesting things from a tech perspective that I’ve had to do to make sure that that satisfies all the full breadth everywhere from, you know, a free version of a simple affiliate plugin, all the way up to the same, with the same solution and the same set of abstractions, all the way up to, you know, a 10 million plus dollar company that has a sales team with different regions. And the sales team has, you know, managers who get commissions based on their sales team’s performance. And then there’s another, you know, they. And all that kind of stuff, stacks and tiers and, you know, everything in between.

Zach Stepek:
You know, if any of our listeners are looking for an affiliate platform, definitely take a look at Siren and you’ll learn why as we talk throughout this episode, just by getting a bit of a glimpse into Alex and the way he thinks. So, Carl, it sounded like you had a question.

Carl Alexander:
No, no, no, I’m good, I’m good. Like, I was just, I keep thinking this was like a recurring topic at Press Conf, like platforming your plugin basically. Um, and I keep thinking about how like, yeah, the, you know, without getting into it, how things have changed in the last few years. That just made it like a requirement as opposed to like something people avoided. Because I, I used to talk about that. I talked when I was talking about serverless stuff like if in Bangkok, like that was 2023, I was like talking about like, why couldn’t Gravity Forms have self hosted and go like, you know, against Typeform for example. Right. Like customers don’t really need to know it’s WordPress underneath. Like it can just be like a lot of plugin businesses are just, they’re platforms, they’re just running inside WordPress. So if you take that outside of it, affiliate’s a good one. You know, like you can have an affiliate pro platform, but it doesn’t have WordPress. Can just be one integration of it. Like it doesn’t have to run inside of it. Right. So even if, even if the platform is WordPress, like, even if the platform is actually WordPress that runs the entire thing, like it doesn’t have to be like, you know, that was like Matt’s dream like 10 years ago when he was like talking about the WordPress OS when he was talking about WordPress OS. Right. Like WordPress as an application platform.

Alex Standiford:
I do think, I do think that, that we are closer to a reality like that than absolutely we’ve ever been in that regard. Like, I’m seeing some things as a, this is a tangent but like that’s something I’ve seen just in some of the things that I’m seeing people contributing and creating. I’m like, man, there’s a lot of stuff that is very much not CMS shaped, especially with MCP servers. And it’s like the identity of what WordPress even looks like over the next decade is fascinating.

Carl Alexander:
Yeah, like I’ve shed a lot of my WordPress like not intentionally, but I’ve just ended up shedding most of my like WordPress consulting and the little I have left. One of them is like literally. So Roots has this thing called Radical, which is like a Laravel WordPress hybrid project type basically and they run their entire licensing platform on top of that. Right. So it’s got some WordPress things, it’s got like I’m doing a lot of stuff to migrate a lot of the internals to be more laravel like but basically. But it’s a platform. Like it’s literally their licensing dashboard. But it’s built in WordPress like, but it could just be Laravel or it could just be WordPress. Like it’s. But again I think the challenge around that I’m sure you understand is like the lack of primitives and stuff to like build stuff on top of, you know, you know, just not having composer support built in, you know, just like, you know, you got to use Roots bedrock or something to like get that stuff in is a good example of like things like that where. But it doesn’t change that a lot of people do, you know. Like what, like either they run their EDD licensing platform through that or like all that stuff. Right.

Alex Standiford:
Like, and what’s, what’s interesting like what you’re talking about where it is a WordPress plugin and it’s also a self hosted service that can be ran independently. Um, in that case the self hosted service is not running WordPress under the scene, behind the scenes at all. Um, it is built using PHP Nomad, which is my framework that actually abstracts. It’s. It’s effectively a series of abstractions for pretty much any kind of common operation that a web based service needs to leverage. And it’s built in a way that it uses a lot of strategy patterns and inversion of control. So it’s using container, like, you know, dependency injection, basically. With that, basically what you can do is you can say you can build your entire application and not even know where what platform or where it’s going to end up. And then after it’s built, you then create the application, the, the platform layer, which is, I call it the different layers. Right. So the platform layer and all that does is it just integrates and implements all of those different strategies, like the database strategy, the REST strategy, the fetch strategy, a cache strategy, stuff like that. And those are defined in an integration layer that then define, like, translates all of those strategies into how it would work. So the way that Siren is built literally is the repository is set up to where runs as a WordPress plugin, but it also runs as a completely separate SaaS with an API and stuff like that. And the difference being one of them uses the WordPress integration with all those things and then the other one is using just a series of PHP Nomad integrations that I’ve created for those individual pieces, obviously, whenever. So it doesn’t even have a platform necessarily. It just kind of is its own thing. Yeah, so that’s, that’s, that’s definitely where, that’s where I’ve gone with a lot of, a lot of that stuff. And it sounds like it’s kind of related to what you were talking about with, with what Roots has been putting in. They’ve been, they’ve been kind of on the forefront of Laravel and WordPress for a long time. So I’m not surprised they’ve gone there.

Carl Alexander:
AI has been a bit helpful for that. They don’t, they really don’t. They kind of like, you know, I, I love them. Like, I’ve been using Bedrock basically for a decade, but the radical stuff and Acorn and stuff is like very much like, it feels almost like straight up what I like, you know, this is a cool thing but like good luck, you know, like, like good luck and, and you know, like I built like security like authentication layer that could like just figure out if you were like Laravel could figure out if you were authenticated with WordPress, like straight up. Like, you know, like all that stuff’s not, it’s not included. Like a lot is not included. Like they’re basically like, hey, we got Laravel to work with WordPress. Have fun. And it’s like we. And a bit of documentation and stuff. So. But you know, with AI and all that and you know AI won’t save you necessarily. Like, this is a good example where AI, like the jagged edges of AI, like, you know, AI wasn’t trained on a lot of, like, hybrid WordPress Laravel code. Like, it’s just like, you know, you, you kind of have to be able to steer it, you know, Like, I’d be like, especially at the start when there’s like, very little, like, foundational work, like, Like, I probably spent a month or two, maybe three, just like building like, foundational stuff, stuff, you know, like how to make the rest. Like they had like, legacy rest API endpoints. But I need, I want to build new laravel API endpoints, like, how to make the two of them work together. Like, all that stuff, like, you know, isn’t done and needs to be, like, foundationally built. And then after that, you know, you. That’s why you built PHP Nomad, right? Like, you want primitives, and then once you’ve built kind of a bunch of primitives, I mean, that’s why, like, symfony components, I think, is like the underdog of, like, all the PHP ecosystem. Like, nobody talks about it. Nobody talks about it. They talk about Laravel and whatever and whatever. But, like, honestly, like, what made all this work is symfony components.

Alex Standiford:
And what’s funny, it was funny, is I’m finding, even with PHP Nomad, every time that I’m working outside of WordPress. Not every time, but often. Most of the time, whenever I’ve created the primitive, I’m like, okay, what do I need to use? Like, what, what is the library out there that can integrate this? Because. Because my positioning with PHP Nomad is usually I’m defining the structure, I’m defining how these things get initialized and set up, but whenever it comes to exactly how it’s done, I don’t. I try to avoid creating like, like, I would never create. Like, I created a PHP Nomad container class that was like a basic PHP on my class. That, that solved that problem for me. I ended up deprecating it because I was like, I don’t know why I’m doing this. This is outside of what I should be doing. And I, I abstracted it and I said, okay, here’s, here’s how. Contain what containers, what is required, especially

Carl Alexander:
this PSR container now.

Alex Standiford:
Exactly. Yeah.

Carl Alexander:
So you can just build on top of that.

Alex Standiford:
Exactly. So. So things like that. But what I was getting at though, is like, all of, almost every time when I’m outside of WordPress, there is a symphony component that will solve the Integration. So like nine times out of ten it’s like, oh yeah, there’s a symphony integration with PHP Nomad. So. So that’s been kind of accidentally becoming. That has accidentally become what a lot of it is.

Carl Alexander:
It’s crazy to me. Like it’s like again like I feel like that’s like always been the issue with like WordPress is that they like voluntarily decided that they didn’t want to deal with, be a part of that. No, it’s not even contributing. Like they don’t get any benefit back. Like, you know, like my only contribution to freaking WordPress core is finding a performance bug. And you know, God bless Ryan McHugh. But like, you know, like we didn’t need HTTP API, we should have had Guzzle in. And like I found like this crazy. When I was building in here, I found this crazy performance issue and it was solving Guzzle. It was just a, you know, because, you know, Ryan’s gone and then we, we end up with this and nobody’s really actively maintaining it or, or checking in on like what the, the actual like, you know, libraries that are being used every day are doing. We’re fixing and then that’s it. Right. So you lose a lot of upstream benefits. Right? To me that was always also a thing, right? Because like WordPress is incredibly self contained and that’s like an advantage probably, I think probably for the adoption and stuff. But without even bringing in Composer support, you could have just, you know, you use. They used internal libraries before. Like they had PI mailer. They’ve had stuff they could have like brought in Guzzle. Like it’s just like. Right, you know, there’s stuff like that.

Alex Standiford:
What’s funny is on the JavaScript side, they certainly aren’t afraid of that. You know what I mean?

Carl Alexander:
No, I know, but that’s like the irony of it. Like, I mean that’s. You just gotta accept it.

Alex Standiford:
Yeah, I know, I know.

Carl Alexander:
You gotta accept that there’s like A. The PHP gets the whole backwards compatibility thing and then JavaScript just breaks everything all the time. And that’s just the reality of things, I think for me anyways. Like, you know, I still think the PHP ecosystem as a whole is like super good. Basically. Like just before the show I saw like Composer, because NPM has been like a crazy freaking vector for supply chain attacks. And then Composer just came in today and they’re like, here’s what we’re doing and yeah, here’s what we’re doing and like, you know, soon it’ll be like 2fa mandatory for all package maintainers and they have like all this stuff. And you know, I mean, I linked this thing in post status and nobody commented on it. But like, you know, I remember talking with Patrick Gallagher of Grid Pain, like when it. They came out with. When anthropic announced like MITOs, right? And I was like, he was, he was. He’s the one. Like, he deserves credit. He was like, he’s. He was hounding everybody at Automatic and he was like, are we in this program? And I don’t think we’re in it because like last week I. That’s what I was like trying to get at on post status is like Symphony released like a ton of. Of CVE patches at once, basically. And they were all found by mitos and they were all like. None of them were false positives, basically.

Zach Stepek:
Yeah, that’s the Project glasswing stuff, right?

Carl Alexander:
Yeah, they were in project so. And that was like, they. I. I’ll give the. The article to Bob so you can put it in the show notes like the, the blog post. But like, you know, they just got one pass at it. Like they were just like Anthropic was like, you know, like, basically send me to it one time and return like this. Like all these bug reports with reproducible scripts and all that. And they were all. I’m constantly thinking like, you know, why is, you know, thinking of Patrick Gallagher? Because Patrick Gallagher was the one that said that the first time. It’s like, where is automatic in all of this? You know, like we’re running 40%. You know, like this. This is a real thing. Like, the AI security stuff is like real. Like, it’s so real. Like, I think people aren’t freaking out enough yet about it, but like, but on the, you know, GitHub got hacked.

Alex Standiford:
All the repos, you know, all the

Carl Alexander:
repos through a VS code extension that got supply chain attack. It was just like. It’s just maddening. Like I’m just.

Alex Standiford:
Yeah.

Carl Alexander:
So anyways, it’s like. It’s the crazy world we live in. The crazy AI world or crazy AI world now.

Zach Stepek:
Yeah, the Project Last Wing stuff. Just to give a little bit of statistic around everything that. That they’ve found. There were 23,019 findings. 1129 were reported directly to maintainers. Then 1900 additional were reviewed by external security firms. 1726 of them were triaged as 90.8% of the 1900 were true positives.

Carl Alexander:
Yeah.

Zach Stepek:
467 findings from that were reported to maintainers for a total of 1,596 findings, 1,451 of those were acknowledged by the open source maintainers, 97 were patched upstream and remediated already and 88 advisories have been published all based on just running it.

Alex Standiford:
One by which is great for open source. It’s also terrible. I feel bad for those maintainers. Oh my gosh, could you imagine that day? But still it’s.

Carl Alexander:
I mean it’s hard because I thought it was a marketing gimmick. I don’t know. Like a few weeks before curl, the curl maintainer was like yeah, we got true glasswing and like it gave us kind of crap except for one and the one was like not a very high score one but again like I think, I think the fact of the matter is that there’s just a lot of, you know, insecure stuff out there. Like I wouldn’t even, you know, I try my best and stuff and I think that speaks to all of it. And also some ecosystem are just like less secure by default. Like sorry JavaScript, like you’re a false fire.

Zach Stepek:
Yeah, it always has been. Always.

Carl Alexander:
Yeah, I know but it’s, it’s. It again. Everything with AI is like rocket fuel, right? Like it kind of comes down to like what we were. A bit of what we were talking pre show right. Like it’s just everything moves so fast. Right. Like I feel like. I feel like it makes the JS, the JavaScript ecosystem, like fastness make it look like a turtle, you know, like how fast things change.

Zach Stepek:
It does. I mean I just got done in the last week setting up a multi agentic team through OpenClaw that has personalities and you know, it’s. They respond in the manner that they’ve been determined to by me. My entire team is based on dungeon crawler Carl characters because why not?

Carl Alexander:
Somebody got me this. It was my birthday and somebody got. I don’t really read fiction anymore but I was told by a friend that, that reads that this book at least is very good. He said that the new one that just came out isn’t as good as the. The first one but. But yeah, it was a troll gift basically. I got dungeon.

Zach Stepek:
I like it. I’m a fan. Obviously I wouldn’t name my AI agents after characters and base their personalities on them if I wasn’t. So there’s a guide that is part of the. The story he’s in introduced in the first book but his name is Morai. So Morai is my primary agent and then the rest of the team design, quality assurance, all of them are named after other DCC characters. So my design, my design agent is Donut because, you know, Princess Donut is always worried about appearance.

Alex Standiford:
That’s funny.

Zach Stepek:
Yeah. And then Quazar is my, is my QA agent, you know, Katya is my accessibility agent.

Carl Alexander:
Yeah, I don’t know any of the characters. I don’t know if I’ll ever know any of the characters. But unless you do a TV show, maybe I’ll watch it when it’s a TV show.

Zach Stepek:
It’s just really cool to me that, that we’re, we’re living in an era now where I can in instill these AI agents with a personality that is based on, you know, these characters from a book series that I love. And yeah, they react in the way that those characters likely would in these situations. So it’s just fun. And I’m really enjoying being able to set them to a task and then watch them collaborate with each other and, you know, they come back with, with a result at the end. But it’s really neat to see that. So some of the projects that I’ve been working on, I’ve been pretty heavily using AI to rapidly build functional prototypes of a SaaS app that I’m working on right now. I’ve been building blocks with AI, now that there are MCPs that can teach these AI tools how to build a block properly.

Alex Standiford:
Doing a lot of that lately too.

Zach Stepek:
Yeah. I mean, building a block with AI once it knows how blocks are supposed to be built is pretty cool.

Alex Standiford:
I actually had a lot of success with that with. I use cloud design just this past week because I, I’m. I’m working on making a course for non developers to show them how to use AI to be able to actually write code without necessarily knowing how to code. There’s obviously some limits there and you know, some things that have to be considered stuff like that, but framing it and just kind of giving them some context. Well, anyway, so as I’m building that page, I built it in, in cloud design first and it actually was able to create a pretty good mockup. And then I have documentation in place that shows how to build those blocks. So what it ends up doing then is it translated them into blocks, it created them, and I actually ended up using some of the, it ended up using some of the libraries, the JavaScript libraries that are built into Gutenberg to help validate this stuff directly without having to open up the editor and stuff like that. So it created this linter that it’s using and all this stuff it’s really cool actually. It was very clever. So that’s been super, super useful. But then whenever it comes to the PHP Nomad, whenever I’m using PHP Nomad for a lot of the backend stuff that I’ve been doing related to making sure like what Carl was talking about earlier, where the, the agent doesn’t have the context in the background and you’re spending a lot of time just putting those pieces together. That’s been one of the biggest things that I’ve noticed about PHP Nomad in the Agentic era. Right. So before I built it, I built it because I wanted to be able to use something on WordPress and also off of WordPress. And then I kind of fell in love with how it works and how it’s set up and, and just wanted to write everything in PHP Nomad. So like I started creating all these integrations. I just created a, an integration for. What’s the name of the. The app that allow. The setup that allows you to run a native php. So I just set it up so that it can do that. So now you can run desktop apps with PHP Nomad. But what’s been really cool about it as I’ve been going is AI, because it’s consistent patterns is starting to be able to understand and match. Oh, this is the right way to do this. This is the right way to do that in PHP Nomad. Because it’s documented, it’s able to see very consistent. You know, there’s this 15 or so different patterns that we use for everything and we’re just going to stick to using those. But then PHP Nomad has a cli. I ended up creating a CLI that actually has the ability to create, to scaffold code using JSON blueprints. So I’m creating all these blueprint libraries that’s allowing me to scaffold all these common patterns and stuff like that with, with JSON and they’re being saved inside of projects. It can be downloaded as packages and then AI is able to train. It’s kind of like you’re building like this self documenting self like the MCP is the code base in a lot of ways. Right. So what it’s ended up doing is it’s like saving me a ton of tokens now because instead of it trying to figure out how to do it, it’s talking in terms of patterns, it’s looking at those blueprints and it’s just using them to build the scaffolding for all these templates and all this stuff it’s been very interesting to see how AI has been able to impact that too.

Carl Alexander:
AI. I mean, maybe we could like shift. We had like an intense AI discussion before we got on that. We were shit, we should have. We should have hit record.

Alex Standiford:
You asked me how I was doing and I said, you know, I’m okay. Just dealing with the bi weekly existential dread that is being a developer in the AI era.

Carl Alexander:
Yeah, and I was just like, I literally just came back from probably. Well, I mean multiple factors, but I definitely played. I had a pretty big burnout and definitely AI was a factor like in it. Like I can’t, like I was listening to you, Zach, talk about like your open claw set up and. And I’m like coming out of this burnout and like it makes me like tired.

Alex Standiford:
Like I don’t want to do that.

Carl Alexander:
Like, yeah, I’m like I don’t not tired but it was just like I’m like noping so hard basically. Like I’m noping so hard listening to that. Like and we were talking before the just to first context for listeners. Like I’m. I’m not, you know, like I don’t know what people expect but like I’m not that. I mean, okay. I don’t know. I’m not that AI pilled by like the standards of like XD everything app. Okay. Like basically, you know, I never, I’m not using solo to like have a. To spawn. To have an MCP to spawn sub agents and, and orchestrating stuff.

Alex Standiford:
You’re using it to multiply yourself as opposed to replacing myself.

Carl Alexander:
Yeah, I’m basically not like, I’m like, I’ve largely been single or dual treading it like most of the time. But I, I started listening to a podcast of somebody like from open Code that did a burnout for a month and stuff. And he was saying, he was like, he’s. He was saying he’s on Twitter now on X and he’s like, he can see like everybody. He calls it AI psychosis. I think it’s I’m. This is like my autistic linguistic thing. Like I’m just like this is not AI psychosis. But he can feel like whatever. Like you know, if you’re paying for multiple codec subscriptions you probably have an issue. Like if you, you know, I like I did it. I managed to burn out on the $20 a month thing. You know, like it’s like it. The, the problem was like, it’s just. It never stopped. Like he talked. The one thing that resonated with me is like, like it’s not that I’m like sleeping with my laptop open. Like some people like, like you read some insane stuff like on. On X. Like. But I resonate like I would wake up in the middle of the night and I’m like, I’d have an idea and I just want to like basically start working on it.

Alex Standiford:
The barrier to entry to do it doesn’t take as much. Right. You don’t have to be in a non super strong right. So I could just start working on something at any time and I’m not realize how much it’s wearing me down and burning me out until I realize that I’m like writing prompts in my head as I’m sleeping and I’m like, oh my God. I have to like think about this.

Carl Alexander:
I listen to a lot of mostly technical and they talk a lot about AI also. And like that also is making me. Nope. A lot now. But Ian Landsman talks a lot about like, how you can, like you can go like a hundred X to nowhere, basically. And for me, I think that was like part of the burnout is like. And I’m seeing it too. Like, I, like, I talked to a lot of devs like, that are outside the WordPress space. Like, you know, like, I have a friend who spent like almost a grand in tokens in a day. In a day? Yeah, in a day. In a day. Like, and he wasn’t like trying to game it or anything. He’s just like running a whole bunch of agents and stuff. And, and. But the cognitive load of that is like insane.

Alex Standiford:
Yeah. And it’s invisible. It doesn’t. I think that’s the biggest thing right there is the cognitive load. People aren’t aware of how much. And I was talking. I touched on this a little bit earlier. I was saying that the idea of. As a programmer, we have for the last 10 to 15. Or for me, the last 10 to 15 years of my career has been even before that I was an, I was a mechanical engineer before I was a programmer. So I’ve always been a technician. I’ve always been the technical work.

Carl Alexander:
Love the mechanics, love the mechanism.

Alex Standiford:
Yes. So I’ve always been the deep work kind of guy. You, you come to me with the hairy problem and you leave me alone for two weeks while I go deep and I, I disappear and I make it happen. I get it figured out and I solve it and I come back and here you go. That. And AI has in a lot of ways inverted that like, like my. It inverts the ink, it changes the inclination, I guess is what I’m trying to get at. So like, there’s downtime where I will have a thoughtful conversation with an AI agent where I’m, I’m working with it, I’m reading what it’s saying, I’m trying to understand and I’m guiding it. And I’ll say, okay, go do that thing. And then it goes. And it churns for 15 minutes, maybe even five minutes. That’s enough time for me to stop that and go. Okay, what am I going to do while I wait for that? Oh, I know. I’ll spin up another agent and think about this other task that I’m working on. And then before you know it, I think Monday. Earlier this week I had. I have a multi. I’m using Linux, I’ve got a multi desktop setup. And, And I had Seven different desktops with seven different agents, having seven different conversations, working on different things that I was trying to take care of at the same time. And I was, don’t get me wrong, I’m super productive. I got all that stuff done. I was very happy with it. And it was absolutely more work than I could get done in one day than I normally can.

Carl Alexander:
There’s no question of the productivity aspect, but I, I don’t think it’s a sustainable seven hours a day productivity.

Alex Standiford:
And I think that’s the caveat that I’m getting into though is that no longer is my work as deep as it used to be and now my brain. I have spent my entire career learning how to work on one thing, go deep, think about nothing but that thing for most of the day, for several hours a day. And that was how I did my best work to. Now AI does not reward that kind of behavior. And it’s been a lot of work to get myself starting to retrain myself to be able to work in a more multitasking kind of way. And I’m not completely sure where the line is, what the balance is between my own personal mental health and my own productivity. Right. So like, and, and, and that, and that bleeds into. Because I’m not working as hard as I’m used to. Like in terms of mentally, I tend to work longer than I normally would because the signal to my body that I’m tired and it’s time to stop for the day is mental fatigue. But my mental fatigue isn’t being itched in the same way it used to be. So it shows up in a very sneaky way that’s different. It’s, it’s like a, a project manager at the end of a day feeling tired because they’ve been in meetings all day. They’re tired because they’ve been in meetings, they’ve been vertical, they’ve been talking, they’ve been, you know, on all day, context

Carl Alexander:
switch a lot more.

Alex Standiford:
Yes. And that versus what it’s like for a technician, pre AI or a person who isn’t using AI. Their exhaustion came from a very different place and it feels different to me. So, so the consequence of that is I’m not very good at identifying, I’m getting better at it, but I haven’t historically been great at identifying those signals. And I think that that was my one way ticket to burnout over the last five months between January and today. I didn’t realize it. I was like, oh man, I could do this all day. I’m not doing any of the hard work. I’m sitting there downstairs, like watching tv and I’m like casually chatting with AI about it, not realizing that I never turned off. And this is just essentially like keeping slack on my phone and I’m just having casual conversations with people at work and stuff like that. And I’m. I’m pulling back on that a lot. You know, I’m starting to set those boundaries and get a feel for that. But. But it, it definitely didn’t just. It wasn’t clear to me at the time that that was happening.

Carl Alexander:
I think it’s still happening. Like, I don’t think this is solved. I honestly don’t think this is solved. Because, look, I was listening to you and I had a similar. By the way, these are all signs that you might be heading towards burnout, FYI. Like, but basically. So the context switch stuff is real. I don’t think I’ve had a flow state since I’ve started using AI. I don’t think anyone I’ve talked to that’s using AI heavily has been in a flow state since they started using AI.

Alex Standiford:
And I miss it. Oh my God. It’s like my warm blanket. I don’t get it anymore.

Carl Alexander:
No, I know. And people lament it and I don’t know if it’s coming back. That’s fine. Like, I’m not, I’m. I don’t think there, like you said, it’s changing the nature of our work. But like, part of the thing that is very difficult is I was listening to you and I went through a similar process where I would. I remember reading somebody that was like, they were just trying to find prompts that would be long so they could watch YouTube or, or go like on X and scroll for a while. And I still catch myself doing that.

Alex Standiford:
How do I keep this thing busy so I can go back to what I was doing?

Carl Alexander:
Yeah, but then you said the thing where you’re like, okay, I’m waiting for it. I guess I’ll just start another one. Right? There’s like, this is not a solved problem. But I’m just identifying what I think is a problem. It’s not solved, but basically we’re stuck waiting. And I don’t think switching to doing multiple agents to do the thing is helping us cognitively. Right. Whether you decide to be productive and make it multiple agents or non productive and just go stroll on social media or right now I just put my laptop aside and I’m playing Diablo and I just go play Diablo while It’s running for 10 minutes and you know, but I don’t think that’s correct. Right. And I don’t think that’s solved unless like inference gets like 100x faster basically.

Alex Standiford:
Yeah.

Carl Alexander:
But they’re all symptoms of the same thing, which is like we are stuck context switching all the time whether you want to or not. Like you have to wait. Like, there’s no question that waiting for 10 minutes for it to do this thing is faster than whatever you would have come up with. But I’ve been trying to do smaller prompts that are faster to do than larger things, you know, so that it. But even then, waiting a minute is still like your brain’s off. Like I’m, you know, like, Zach has ADHD, I have ADHD. I don’t know if you have ADHD. We all have ADHD. It’s like we need that flow state. That’s the only way we’re going to get it. And we need that flow state. You’re like telling you like, just hang tight for a minute. Your brain’s like, nah, man, there’s a fucking squirrel over there. I got to fucking go get, go chase that squirrel.

Alex Standiford:
I’m not built for that, man. Yeah, exactly. Yeah.

Carl Alexander:
And a lot of people aren’t. Right. And that’s like taxing, you know, like

Alex Standiford:
it’s more exhausting than anything. I get exhausted trying to think of what I need to do to make sure that I’m making good use of my time in the meantime.

Carl Alexander:
But that’s a sign, like, that’s what I’m saying. Like to me, like, that’s like a sign that like it’s just there’s. I don’t have a solution to it because like I, I don’t know what to do. It like the.

Alex Standiford:
I don’t even think it’s okay. I don’t even think it has one solution because even from a day to day perspective, some days it is completely okay that I’m spinning seven things if I’m building. Like for example, like, like I. That example that I had on Monday was building individual patterns for a well defined project for a website build that I’m doing. They were patterns that I’m building in WordPress already defined. I already have the process down pat. I already did the work to figure out how to do all that. Right? So it was literally spinning these seven up, get them moving so that the process can do its thing. It takes a while for em to turn nine times outta ten. It’s completely done when it’s done. I don’t have. It’s not an extensive review process. I’m not talking about architecture. I’m not really. I’ve already made the decisions at that point that it’s perfectly fine. For the love of God, take that away from me. Make that happen. Spin those plates. I will get so much more done in that day as a result of that. And I can potentially find other things to do in the meantime, such as playing Diablo 2, which is the best of the Diablos by the way then.

Carl Alexander:
But I was never a big Diablo 2 fan like. But I like, you know, I. I played a lot of it but I had friends that were like. Had bots running like Mephisto bots like, like running the. With teleport source and just like fuck. Just spamming till they get their perfect eagle horn and shit for their all and stuff. It was just insane stuff.

Alex Standiford:
I still play it all the time like almost.

Carl Alexander:
I mean they came out with a new class, they’re maintaining it. Like I love it, it’s a great.

Alex Standiford:
But the, the other piece of that is then there’s also some days where I like today, for example, all of my work is very much architectural. I’m having conversations that are like bigger picture things like. Like I said, I’m trying to figure out this downline structure. So it’s very much one AI chat that I’m having with this one specific thing. And it does feel flow like it’s. That’s about as close to flow state as I’ve ever come. And I think the biggest trick is

Carl Alexander:
still that stuck waiting for a long time. I don’t know. I don’t know what you’re using. Like I’m a. I’m a 5.5x high person so it. And I’m not using slash fast. Maybe I should try slash fast. But it’s like something I’m doing that

Alex Standiford:
helps a little bit is I will. I will preload like just. Just Claude AI. Like the. The app. Not. Not the code, not Claude code or anything like that. I’ll load it on my phone, I’ll put earbuds in and I will have it load the context and then I will conversation. I will have a conversation with the AI instead of trying to solve it at my desk. That’s been really helpful because. Feels productive because I’m walking or I’m doing dishes or I’m doing something while I’m waiting for it. But it’s not something that’s requiring my brain and it’s like it’s letting me kind of distract Myself kind of passively. And then whenever I’m done with that, that information all gets put together in a report that I can then take back to the AI and say, okay, go do this. I’m warmed up, I know what we need to do. Go do it. That’s been pretty transformative for me actually, because it scratches a lot of itches at the same time. Keeps me busy, keeps me busy during those down times. It gets my body moving so then my brain is actually working better and I’m actually able to think a little bit more. I find myself getting in flow state away from my desk more than I do at my desk right now. It’s kind of crazy.

Carl Alexander:
I actually don’t mind it for big architectural stuff, like, but it’s just when it comes to implementing the big architectural stuff, it just, there’s a lot of back and forth. Like, there’s a lot of review. Like, I don’t, you know, like, yeah, you have Aaron Francis and stuff that like, doesn’t look at code and stuff. But, like, I don’t, I don’t see it. Like, I haven’t been able to get it to like, you know, take off my hands off the wheel kind of thing. So you’re there, you’re reviewing. I don’t have a good setup yet for like doing all those things from one screen. Also, like, I just, you know, I think IDs and stuff. Like, I still want to go click and code. Like, I still want to be in PHP Storm sometimes. Still want to see the git stuff separately. I still want to chat, I want everything in a big window because I work from my laptop. But yeah, I think there’s something there. Like, basically, like, I don’t know, like, I feel like everybody I talk to that uses AI a lot is like on the verge of burnout. And I think the context switching is like one of them.

Alex Standiford:
Yeah. For me, my process.

Carl Alexander:
Yeah. One thing we talked about too, that you kind of alluded to is like, and I’ve talked about it with somebody else is like, you can go a hundred ten, one hundred x to the wrong direction, but also like, I’ve. I have no boundaries. Like, you are a business owner. I’m a business owner. Like, you’re trying to build a SaaS. Like, there’s no boundaries. Like, I’m going to rewrite the emir runtime in Rust at some point and it would have taken me months to do it. I would have been able to do it on my own, I’m sure, but it’s going to get Done. But there’s so many of these things, right? Like there’s so many of these things that as a business owner like the price of entry was like basically too high to consider and now like you have limitless things of triage to do.

Alex Standiford:
Yeah, I just rebuilt Siren’s entire front end end to end in React because it was built, it was like darn near prototyped for WordPress and it was like, you know what, it’s time, I’m going to do this. Let’s go.

Carl Alexander:
Yeah, that’s what I have to do soon as to redo the marketing. That’s also part of how I got to burnout basically because we, you, you’re like oh like let’s do this thing. But there’s no question it’s better. Like Emir has to manage. I think I’m close to 10 php versions. A lot of them have to be built by hand. Like I had to like migrate like the Amazon Linux version. It would have taken me months if I have even been able to because I don’t know compilers enough to debug compiler errors which are inscrutable to somebody that doesn’t work with GCC and stuff like that. It just knows them. It knows it. It did it in a day, it would have taken me months. And it was just like there’s no question. So you, you have limitless possibilities, no boundary. Like that’s kind of like my thing out of Burnout that I’m trying to do is like more boundaries and I

Alex Standiford:
think that kind of touches to like what I was in the pre recording chat, we were talking a little bit about it and I think that a lot of the other piece of the burnout is so, so one obviously the, the insidious thing that I was talking about. But the other thing that I identified whenever I’m thinking about it is also as developers we are inherently curious people. We are tinkerers. The reason why most of us are this are where we are is because we want to know how things work, how far we can push things, what we can do with them, what capabilities is it unlocking?

Carl Alexander:
I mean that’s why I like AI. Like I’m not having an existential. Because there’s two camps around AI, right? Like there’s people that like view that viewed their code as like something beautiful that he wrote like kind of like the DHH camp, which is ironic because DHH uses a lot of AI, but. But DHH was always about po. Like the coding has to be beautiful and that’s why he loves Ruby and stuff like that. Primagen’s kind of like. And look, I. I started coding when I was like 6 and like, I just like building stuff. For me, it’s like Legos. Like my favorite analogy of programming is it’s like Legos. I get to build whatever I want to build in the digital world. It’s fucking awesome.

Alex Standiford:
Yeah.

Carl Alexander:
Like AI is just literally the 3D printer. It’s like a genie. Yeah, it’s like a genie. Like you know you want it, it’ll build it, it’ll help you get there. Right? Like it’ll. It might not do it well or whatever. Like it’s kind of like the, it’s like the monkey paw in Simpsons, you know? Like it’s the monkey paw. It’s the monkey paw in the Simpsons. Right? Like it’s like you get your wish, but it’s like sometimes it comes like with, with a side effect, you know, if you didn’t do the wish properly.

Alex Standiford:
And I think that related to that. The curiosity. So the curiosity just made. I know that’s what it. One of the big things it did to me as soon as I saw. I think it happened in like December for me really is wherever I saw cloud code, Codex and all this stuff really taken off and I was like, oh my gosh.

Carl Alexander:
I talked about in my year review. I think like for me it was Gemini 3 started being really good and then Codex came out around that time. But I was just on the Google train at the time. And then Opus whatever came out too and everybody started being like, okay, this is legit now.

Alex Standiford:
Yeah. So once that happened, I really started. I mean, I was using AI before, but that was whenever I started really using it in the terminal context and stuff. And that just led to a flow flurry of activity for me. I built. I’ve probably built five or six different PHP Nomad services, like an internal note taking, second brain kind of thing that I’m sure every other developer has already built over the last five months too.

Carl Alexander:
But I did. I have no second brain. That’s why I’m like, I’m like shocked. Like, I’m like, I. I don’t think, you know, like, I think you can get there. Maybe it’s like just my crazy ADHD and stuff and like we were talking like the winter that never ended in Canada, but it was just like all

Alex Standiford:
of those things like culminate into burnout. Right. I’m going to spend all my time trying to understand this. I’m building a lot of stuff just because I can And I’ve never done it before. I’m trying to learn what this thing can do. I’m working in a different mode of work where I’m multitasking. Also. I’m still trying to do my job and figure out where this fits into my career. And like, all of this, it’s just like.

Carl Alexander:
And there’s this manic energy.

Alex Standiford:
Yes, manic is a great word for it.

Carl Alexander:
Like, it’s receding now because I think people are hitting burnout. And like, also companies are realizing that they can’t sustain, like, how much they’re spending on AI internally.

Alex Standiford:
I also think that there’s, like, probably some quality issues they’re probably not talking about. You know what I mean? That.

Zach Stepek:
Oh, absolutely.

Carl Alexander:
I think some of the outages, like, from. Well, some are known unofficially, like, like, AWS had an outage that’s like, apparently like Kiro, like, basically overrode some production shit. And then. And then Cloudflare also moves really fast. Like, I think some Cloudflare outages have to be like, AI generated. Like, it’s not officially out, but yeah, I mean, that scares me a lot. Right? Like, that’s why I’m like, emir is an infrastructure product. Like, I take this shit seriously. Like, you know, it doesn’t change that it’s really good. I think the most underrated feature for me of AI is I can just be. Because I have to context switch a lot. Like, I. My dream has always been to work on the mirror full time so I wouldn’t have to context switch between consulting and all these other things. But I think one of the unintended benefits of AI is I can just be like, hey, it’s been two weeks, I haven’t worked on this. Like, can you, like, get me up to speed on where we were? And like, that could take me half a day to a day before. Like, I was just like, try to remember what I was thinking and stuff.

Alex Standiford:
And again, that’s a great example of what I mean about when I was talking about walking. So I’ll do that while I’m going on a walk because it’ll be like, I’ll be like, warm me up. I want to get back to my desk and start working on this. That I can’t remember. I do that before meetings too. I do that before all kinds of stuff. Like, it’s so useful.

Carl Alexander:
Yeah. So it’s good for that. Like, I love it for that.

Alex Standiford:
I do definitely feel like I’m around my desk a lot less, which I think is just generally good for my Health, too.

Carl Alexander:
I mean, I’ve read that from a lot of people. I mean, this is like, maybe a hot take. I don’t know. Like, everybody is like, like I said, so manic. But I think you should use AI to reclaim your time right now.

Alex Standiford:
A hundred percent.

Carl Alexander:
I think the productivity gain is there. But, like, this is an interesting podcast that I didn’t finish that. It’s another person related to Prime Edge. And it’s like Casey Monte Montessori. He interviews a AI guy that’s been in it for, like, 20 years, and they talk about the productivity. But the thing is, nobody actually knows how much, like, productivity this is giving us. Like, we know we’re faster, right? We’re faster at doing stuff, but are we faster at doing stuff that makes more money for the company? Like, even for you, for Siren and stuff like that. Right. For our business? So, like, you could just do stuff that makes you more money faster. And it’s just. That’s it. You’re done. Like, you just take the rest of the day off, spend time with the kids, go touch grass, like, you know,

Alex Standiford:
and that comes back to what you’re talking about with boundaries, right? So that’s. And that’s always been true, but, like, that’s the.

Carl Alexander:
So this is what I’m trying to do coming out of this burnout, basically. Yeah. I’m like, I’m trying to chill. I’m trying to not be like, again. There’s a lot of factors at play. Like, it’s not just AI, but what AI does is related because it’s hard to put boundaries when AI removes boundaries.

Alex Standiford:
Right.

Carl Alexander:
So you have to be proactive. It’s like, kind of like Zach’s like, I’m always so impressed, but, like, Zach’s on a weight loss journey and.

Zach Stepek:
£91.

Carl Alexander:
£91. Wow. Congrats, dudes. Yeah, but, like, one of the classic things of like, of. Of food is, like, don’t have in the house. Right? Like. Or stuff like that. Right. But, like, the boundaries before, where you’re like, okay, it’s going to take me a month to do this, or stuff like that. Like, now your boundaries are gone, so you have to do a lot more self work to set those boundaries. And that’s hard, right? It’s hard when you’re having all this external pressure of, like, for me, it was like, I support Laravel. The marketing doesn’t talk about it yet, but I was like, trying to go after, like, vapor, that sunsetting. And then at some point I was just like, none of this like, actually matters at the end of the day. Like, it’s just. And I’m just like, rag. I’m making myself, like, sick, you know, so now I’m like, okay, like, I work, but I’m more. I’m trying to go back to where I was before, which is like a bit more casual, you know, set some boundaries and stuff. But it’s hard. It’s hard because again, there’s this manic energy. We don’t know what our career. You know, you started the podcast. You’re like existential dread about AI. Like, nobody knows where anything goes. So you feel like you have to be running. But we don’t know where we’re running to.

Zach Stepek:
Yeah, no, absolutely.

Carl Alexander:
We feel like we’re running, so we run. But we’re like, running like headless chickens. Like, I say, like, nobody really knows. Like, I’m not convinced that AI, you know, the Sam Altman and the Darios know exactly where this all ends up. Right.

Alex Standiford:
So how could you.

Carl Alexander:
No, exactly. So we’re just all kind of freaking out. And like we said, like, everything changes so fast. Like, even the sentiment towards AI coding is always fluctuating. Like, we’re not at the beginning of the year where it was like, okay, we’re, you know, like, you gotta gas town it. Sorry, I just swore. But like, I was just like, you got a gas town with like, you know, the, you know, let it rip and all that. Like, now, like, this, the CFOs are like, wait, everybody spends like a thousand a day on tokens. Like, we can’t do that.

Alex Standiford:
There’s going to be a lot of conversations a decade from now that we’re going to look at back to this time and be like, oh, my gosh, that was crazy. I can’t believe we actually did that.

Carl Alexander:
I believe it, Alex. Like, I cannot, like, but again, it’s. We’re all striving to figure out, like, we’re all trying to stay afloat and also figure out where this is going and all that. So it’s, it’s all legitimate. Like, it’s not like, you know, but it’s. How do we do that in a healthy way?

Alex Standiford:
Yeah. The good news is, though, this isn’t the first time that we as a society, or even we as developers have had to experience things like this. Maybe us specifically, I don’t know. But, like, I’ve. I’ve. I read a book called I, I, I’m a big fan of the, The Getting Things done methodology. So that book and that, that, well, he. They made one recently for teams, so getting things done for teams and it’s a more up to date version of it. And, and I was reading that book and he’s talking a lot about some concepts like Holacracy and stuff like that. But he was reflecting on how company structures change. Like entire company structures transformed almost overnight as a result of email because email communication existed and the way that it changed everything, it changed the pace of work, it changed how things could be done and it changed entire company structures. And that’s even one example that I’m thinking about. Like you look at how companies are structured today, Agile workflows and stuff like that, they kind of work, they can work with AI, but there’s an entire potential, there’s entire company structures we probably haven’t even full. We definitely haven’t explored yet that look nothing like what we even know now because AI exists.

Carl Alexander:
Yeah, I still think it’s happening a lot faster. Even if those were fast. Like I still think this is happening like, like very, very fast.

Alex Standiford:
Absolutely. But the thing is, is eventually it’s going to have to like there, there’s going to be a wall that’s going to get. I think that’s going to hit there because you run into. Because if you start where you talk about company structures and you’re talking about things like that they can as a company only move so fast.

Carl Alexander:
You know, I mean I talk to a lot. If you look at the, the layoff stuff and talking to people that I. It’s all middle management stuff that’s being all lowed out. Right. Because they’re just like, we don’t need as much of it because like the middle managers are the programmers now basically. So. Yeah, but that doesn’t mean that you won’t need middle managers in some capacity. But like I think that like that’s where you’re, what you’re saying is like, I think companies are being rearchitected right now. Like I think it’s already started for that for sure.

Alex Standiford:
It’s all untested.

Carl Alexander:
I, I was talking about that stuff though, like cognitive stuff at the, even my year in review or I wanted to. Like, I just, yeah, it’s just, it’s very interesting. Like I just, I don’t, I don’t know where this ends up. Like that’s why like I’m just like, I think the best thing we can do right now is like try to reclaim our time. Because like I said, like, I don’t. You know, I remember at Press Conf there was Some discussions where like, you know, there was like some people that were like expecting developers to be like 10x more productive. That’s a lot. No, I don’t think you can realistic like be 10x more productive.

Alex Standiford:
I was already teetering on the verge of burnout before AI. So like I’m with you 100%, Carl. Like I’m, I’m taking my time back.

Carl Alexander:
Yeah, I mean, I think you should take your time back. I think you’re still more productive even taking with your time.

Alex Standiford:
Oh yeah, no doubt about it.

Carl Alexander:
You know, there’s no question. Like for example, the Amazon, the Linux upgrade stuff, there’s stuff that it’s going to be more than 10x, but again if it’s 10x but it doesn’t affect your revenue, it’s not 10x. Right.

Alex Standiford:
It’s 10x.

Carl Alexander:
You got there 10x faster but you’re not growing your revenue 10x. Right. Like it’s. So you have to either do with less people, which is a bit of what’s happening.

Alex Standiford:
Well, I also think that another thing there is the time that you’re gaining back needs to be spent on being use if you’re using AI or not being more thoughtful about the direction you’re heading. Right.

Carl Alexander:
Yeah. The triage is more important than ever.

Alex Standiford:
I’m spending a lot more time understanding that than ever.

Carl Alexander:
Absolutely.

Alex Standiford:
Usually I. Because usually my inclination, whatever I’ve did coding in the past was if I was just doing it by myself, I didn’t have a team, I didn’t have anybody to report to, and I was doing it my way, the way I want to do it. My comfort zone was I’m going to start building this and I’m going to kind of just figure it out as I go. Right. There’s going to be some, there’s going to be some discovery. I’m probably going to do some big picture stuff, but I’m not like, I’m not having conversations or thinking about the big picture of everything that’s gone now. Like, if I try to do that same process with AI, I with, with something as, as, you know, programmatically complex as something like Siren or something like that, it falls on its face hard because it, it just, it will try to do that and figure it out as it goes and I will spin my wheels and have to, you know, like, clean up the feed, clean up the conversation, restart stuff. Like that’s way more frustrating. So like, and that’s just in the programming context, that’s not even talking about the context of like deciding If I should be doing the thing that I’m tempted to be doing right now in the first place, you know, that’s why I was talking about like that second brain idea and stuff like that or concept that I’m using stuff like that because I’m doing, I’m doing a lot of capture. I’m capturing a lot of my, my thoughts and stuff like that. I’m. I’m using AI to do a lot of journaling where I’ll talk to it and I’ll say, okay, this is great, make this a journal entry and save it in this place for me so that we can come back to it later whenever we’re conversing and it’ll like reference that stuff and actually I got my ADHD diagnosis this year because of that. I was journaling and talking about my frustration about being so like all over the place in terms of like, I can’t, I’m having more trouble than ever focusing and that like, and all these other things and the conversation it led to like me just like really frustrated about it. And it was like, Alex, I’m looking at these things. I’m looking at these patterns. These patterns are a sign that you might want to go get an evaluation on something.

Carl Alexander:
Right.

Alex Standiford:
I don’t know what it is, but you might want to go get an evaluation. And I kind of already knew, but I did. And it, you know, it was just like, I don’t know, man. That’s one of those things where it’s just like the, the low key, like transformative thing that, that hasn’t, that isn’t being talked about.

Carl Alexander:
I wrote a whole thing on that in my year in review about using it like on top of therapy and to like figure something off and, and stuff. Yeah, it’s just in the context of like work where I’m like, I don’t, I just, I don’t know where the, like, it’s shifting the balance and it’s like, where do we end up? Like, like one of the things that I’m like not sure is like the seven hour a day, you know, how sustainable it is with this.

Alex Standiford:
It, it never has been.

Carl Alexander:
No, it’s never been.

Alex Standiford:
But I’ve never been able to work for eight hours a day. I’ve been able to barely.

Carl Alexander:
But there’s, but there’s corporate expect. We’re all self employed here. So like I’m just, I’m just, I’m. I’m thinking for my friends and stuff that are in a corporate environment where there’s expectations around AI but there’s also expectations about, like, you know, how much you should work per day and stuff. And I’m, you know, I mean, some of it is just. There’s a lot of meetings and stuff maybe, but I, you know, it is very demanding.

Alex Standiford:
Yeah.

Carl Alexander:
And we lost on top of the flow state. We lost a lot of like, just easy work.

Alex Standiford:
Yeah.

Carl Alexander:
That we could, like, just shut our brain off and do.

Alex Standiford:
Yeah, you’re right. You’re right.

Carl Alexander:
And we’re like, AI Kind of left us all with only the hard stuff. Stuff to do. Right.

Alex Standiford:
That’s really true. That’s a great point.

Carl Alexander:
Yeah. And like I said, like, I’m just. I’m just always an observer of this stuff, but I. I’ve always been fascinating from, like, when I applied for my MBA about like, work culture and like the workplace and stuff. And I. I just keep thinking about this. This idea of like the seven hour a day doing this really high cognitive work. And I’m just like, where there’s no more kind of like breaks and you know, like. And then you read people that are like prompting their agent while they’re in a meeting or right before a meeting, and it’s just like.

Alex Standiford:
Yeah.

Carl Alexander:
That’s why I’m like, I. The burnout thing, I think is gonna show up more and more is my educated guess, like, because I just don’t think it’s. We’re in a sustainable phase right now. And there’s like the whole dopamine aspect. We. We kind of talked a bit about it too before the. Oh, my God, Bob’s gonna kill us. But it’s so long, this episode. But. But the dopamine, the dopamine aspect of like, AI too. Like, it’s like slot. Right. Like.

Zach Stepek:
Yeah, it really is.

Carl Alexander:
So that’s real too. Right. So there’s just all these factors where I’m like, I’m. I was very curious even at the end of the year and as the year was starting, like, where this is going to end up. But, like, like, I thought I was good. Like, I literally thought I was good. And then I. I’m like collapsed. And like I said, like, I was just, like, I don’t feel I was like the most AI pilled of all. Like, you know, like, I’m not Zach with an open claw team or you who’s running a bunch of agents. Like, I was running at most two or three.

Alex Standiford:
Yeah.

Carl Alexander:
Even then my brain, like, I talked a lot about it in therapy and I think we can. All three of us can relate to this. It. We Couldn’t. I can’t. Couldn’t shut my brain off.

Alex Standiford:
Yeah, that’s what I’m saying. I. I’m not kidding. I wasn’t joking when I said I was literally dreaming in prompts. Like, like when I was at my worst, like, I was like, holy crap, this is bad.

Carl Alexander:
This can be an entire fucking funny. I don’t actually have dreams, so. Yeah, I think it’s a blessing, actually.

Alex Standiford:
Yeah. But like, I would wake up and I would be like, oh my God, did I actually write that in? AI. Like, I would like, think that I was talking to, like, it was. It’s insane. That was the weirdest.

Carl Alexander:
But I would just wake up and I. I’d have an idea or I’d be like, oh, I want to keep working on it because, like, I’m having so much fun. And the same way slot machines can be fun.

Alex Standiford:
Yeah.

Carl Alexander:
Right, Right. Like the same way playing poker can be fun. It’s. It.

Alex Standiford:
The only difference is, is now sometimes this stuff actually gets done.

Carl Alexander:
Absolutely. And it gets done better. Like 5.5 is a lot better and stuff. Like, I get it, but.

Alex Standiford:
But that also kind of makes it more dangerous because I know it can get done in a night.

Carl Alexander:
Yeah.

Alex Standiford:
Do you know what I mean?

Carl Alexander:
Yeah. But that’s why the boundary setting, I think, is super important. I think, like, being able to do that is so key.

Alex Standiford:
On the topic of the seven hour, like the work week and stuff like that, I have always doggedly fought against the notion that the office job should be in the hour. I’ve been fighting that since 2010. Right. I’ve always thought that was dumb. That’s the whole reason why I work from home. That’s the whole reason why I work for myself, all that stuff.

Carl Alexander:
But like, yeah, same.

Alex Standiford:
I’ve never. And I. And honestly, I don’t even believe that most people who are working in an office eight hours a day are truly working for eight hours.

Carl Alexander:
But the funny thing is, I probably worked more than seven hours a day, like during this phase. Right. Like, I, like, I was working like legit, like seven days a week.

Alex Standiford:
Yeah. 6:00am to 10:00pm like, yeah, I. I

Carl Alexander:
was like, I enjoyed it more than playing video games. I was not playing video games. I was not doing anything. I was going to the gym, I was eating and I was coding. I was doing those three things for five months, basically. And I hated life because it was fucking winter in Canada and it was like the worst winter I’ve. Everybody is like, one of the worst winter is like in my entire life. Basically, that’s all I had to do. And that. And my brain wouldn’t let me do anything else. Like, it wasn’t even, like, I was, like, not enjoying it or like. But it was just. But there’s like this. Like I said, it’s. It’s. There’s something, right? Like, there’s a. There’s an addiction. There’s a. There’s a dopamine. There’s like. There’s the fun of being limitless. Like, again, I think you can. I’m sure you could say that, Alex. Like, if I wanted to, I could work seven days a week, 14 hours a day for the next year, and I would have stuff that I want to do.

Alex Standiford:
Yeah.

Carl Alexander:
For my business.

Alex Standiford:
Yeah. That’d be true with or without AI but it would be more true.

Carl Alexander:
It’d be true without. But even. But it’s crazier with AI, Right? It’s crazier with AI it was definitely true without AI but before AI you would be like, yeah, okay, but I can’t, like. Or I just don’t want to spend a month, like, you know, I wouldn’t want to spend three months learning rust and writing my runtime in rust.

Alex Standiford:
But here’s the interesting thing is, like, I. I knew then. This is what I mean about. About limits, right? I knew then, before AI, where my burnout point was. I knew what the signs were. I knew how far I could push, but I knew I could do, like, a really hard sprint working on nothing but a coding project for like a month, maybe six weeks. And after that, I knew that there was a cost. I would be burnout and I would have to have a few weeks.

Carl Alexander:
I feel like you’re like me. I talk about it in therapy.

Alex Standiford:
Yeah.

Carl Alexander:
I’m like Icarus. I’m like Icarus. I’m like, how close to the sun can I fly?

Alex Standiford:
I have to know and I have to. And that’s what happened here at the beginning of this year. Absolutely. So, like, so what I’m finding. Okay, now I know AI, it took about double the amount of time before I was. I was in that same state. And it’s like, okay, but. But it was worse in so many other ways. And there’s other consequences of it, of a. Of a thing like that going longer. Right. Like, like a four to six week period. I can put it on my calendar and I can say, okay, that is like post Christmas to mid February, nobody’s doing anything anyway. I can, like, I can do something, recover and then be ready for spring. But like. Like, A longer sprint like that with a. It’s just. I don’t know, it. It’s limits just learning those things. And I think that that’s been a lot of what for me. That’s why I keep coming back to.

Carl Alexander:
I know I’m so well acquainted. I joked that man burnout or like very good friends, you know, Like, I. Like. Like I talked about it before Press conf. But I was so sure I dodged it. I mean, it was a bunch of factors. Like, I. It was hell to come back from press conf. I got like, stuck on planes, stuck in Atlanta, stuck there, you know, but my therapist was like, yo. You were talking about how close you were, like, right before you left, you know, so, like, you didn’t gouge things. That’s the thing about being Icarus. You know, you get the numbers off and then, poof, you fall back down. So. But yeah, no, it’s like, it’s. Yeah. So I think we should, like, wrap this up for poor Bob, probably. But yeah. It’s so good talking with you, Alex. And as Zach always says, where can we find you?

Alex Standiford:
Yeah, so you can find me at alexanderford.com. um, that’s the primary place where all my content and stuff is published. You can also, of course, find me on x. I’m on LinkedIn.

Carl Alexander:
I’m.

Alex Standiford:
I’m cross posting most of the content I publish on there as well. Um, you can find me on Mastodon directly if you go to my website. You’ll be able to connect with me there. Um, those are. Those are the primary ways. Of course. You can always check out Siren at Siren affiliates. Com. That’s where I’m doing most of my work. Most of my energy has been going into that.

Zach Stepek:
Right.

Carl Alexander:
All right. Always great having you on, Alex.

Alex Standiford:
Yeah, pleasure is all mine. Thanks for having me.

Zach Stepek:
Anytime.