Nate Stewart joins Luke and Jonathan to discuss chickens, gambling, commercialized addiction, skill atrophy, taste, GeoCities, and how Automattic, WP Engine, and the WordPress community might work together in a way that would make everyone happy. Nate share his perspective on an unsuccessful retirement and advocates for more organizational diversity in the AI ecosystem. Nate also checks off his bingo card while Jonathan and Luke realize, again, that they’re getting old.

Transcript#

Luke:
Nate Stewart. How many chooks do you have? Do you call them chooks, or is that an Australian thing?

Nate:
We don’t — only in Australia. We call them chickens. They’re part of the flock. We have, I think, twenty-seven. My wife’s gonna get mad at me — if we have twenty-eight and I said twenty-seven, it’s like the end of me. But it’s ever-growing. So actually, backstory there: we flew the chickens we had in Puerto Rico to here.

Luke:
Wow.

Nate:
We had about twenty, and then some passed away when we got here. But there were a lot of chickens around the area that were being abandoned, we found out. In parks — and there’s like a local Instagram thing —

Luke:
Roosters or chickens?

Nate:
Roosters and hens. Roosters actually have a pretty big problem with being abandoned in general.

Luke:
Yeah, yeah — I want to abandon a couple of ours, let me tell you. I’ll ship them. I’ll ship them over to you.

Nate:
Dude. Bring them over here — we’ll take them. So we started getting into that. My wife rescues them, and we’ve rescued a couple of hens and roosters. So we’ve added to the flock quite a bit since we’ve been here.

Luke:
Do you have an incubator?

Nate:
We don’t. I mean, look, if left to Jess’s own devices, we’d be incubating — we’d have like four hundred new hens and roosters per month. But actually, shout out to my wife. Tropical Chicky is her Instagram. So if you go to TropicalChicky — and tropicalchicky.com, which is a WordPress site — all of this stuff’s documented on her Instagram.

Jonathan:
It’s pretty incredible. I’ve grown a newfound — my wife is also friends with his, and they’ll talk about chickens, and we’ll be traveling the world and we’ll see a chicken thing and Amanda will go, hey, Jess would like that. And you not only talk the talk, you walk the talk. A chicken would be very lucky to be in the Stewart household.

Nate:
One hundred percent. Yeah, they live better than I.

Luke:
No — if the chickens get inside, Jonathan, then they’re getting in lots of trouble, let me tell you. Not in the household — just out in the coop, out the back. You don’t let chickens in the house, do you, Nate?

Jonathan:
I know — they bring them in the house.

Nate:
Oh no, we a hundred percent do. In fact, later — so there’s a chicken that laid down and might be a little bit, you know, unwell. Jess was telling me right before I jumped on here, we’re gonna bring in Pecky — that’s her name — and she’s gonna get some house time later. A little spa, probably a little bath, treats, all that.

Luke:
So Jonathan, you were saying earlier that you had an actual intro for Nate. Were you going to ask him a real question?

Jonathan:
Nate, I’ve got a question for you. With no prior context to this, how do you feel about the recent announcement that BigCommerce and WP Engine are in a partnership together? It’s now public news.

Nate:
Oh man. Look — Josepha, I listened to that episode, and I can actually say a very similar statement there, and it’s honest. I’m about a year removed from being deep in with Commerce, and although I led strategy there and kind of knew all the WordPress ins and outs, for a while there it really was fighting to get more involved in the WordPress community like we originally had. And it was like, what are we doing? So on one hand, I think it’s good that they’re trying to invest in WordPress. Because I know the data — the backing data is thousands and thousands of merchants on Commerce, and beyond on different platforms, use WordPress, and having them not have to move off to get the capabilities of the other platforms they use — great, right? Now, why they wouldn’t go deep into the community in general, and not just with WP Engine, I don’t know the ins and outs of that decision. So I think it would be great — and maybe this is just really hard to do currently, maybe we could talk about that — but can you have a company and have a really specific, well-thought-out, executed strategy that has Automattic, WP Engine, and the community all happy? Is that possible now? So I’ll throw it back to you: if we’re giving free strategic advice to not only Commerce, but a company going through that — you’ve got WP Engine on one side, you’ve got Automattic and the community — what do you do so everyone is, let’s say, eighty percent happy?

Jonathan:
The first thing that comes to mind for me is you have to decide whether or not you ignore — or how you think about — the legal battles that are ongoing. And I know that some folks in practice just ignore it. It’s like, all right, well, life goes on, things need to be getting done. These are big companies we’re talking about as well, and so there’s always more going on. And setting aside WP Engine and Automattic specifically, this is just a general truth about big companies: it’s not uncommon to have folks who don’t talk to each other, or to have this thing over here happening that seems entirely separate from this thing over here. And sometimes they are, and sometimes they’re not. It’s like, well, we’re gonna try multiple things at the same time and see what happens.

Luke:
I think you’ve also got to have that dedicated person who is the community person on staff, you know, and you don’t take sides. If you take sides, you’re going to lose one of the two sides. You just focus on WordPress and don’t focus on the fight. I think Elementor is a great example of someone who’s doing a really good job of that, by the way.

Jonathan:
Yeah. So Nate, you have a decade of prior context at BigCommerce — it’s Commerce.com now, but I still know it in my head as BigCommerce; it’s one of the companies in the portfolio. Before that, you had a startup that you grew. You’ve had a long background in tech. You’re not doing any of that now, and of all things, you’ve decided that you were gonna focus on this world of AI. What drew you to that?

Nate:
Yeah, I mean, it was a bit of a journey. I think all things rooted in worth doing are kind of not straight-line. They’re more of, what are other people having problems with? I set out to just take a very long break and chill in Puerto Rico at the time. And — we were talking about Jess earlier — it was kind of, hey, I’m going to do that. We picked a date in 2025 where, like, this is the longest you will stay at BigCommerce. And so I did all the planning and whatnot, and it’s like, okay, cool — I’m ready to be me and my wife and my animals and just chill. And I tried that, all right? Let’s go on the record that I did try that. I chilled, we went to the beach, I did some permaculture stuff. And I was dabbling around in tech, and at that moment, vibe coding, but it was more for fun. It was more of, there’s this trend happening. And at the time, everyone was nearly a hundred percent against it. And it was a growing trend of, okay, there’s Lovable, there’s Bolt, all these things — that’s before Claude Code and a lot of the other terminal-based things kicked off. So I’m really just having fun of, like, what is this? Right? Not judging it, just trying to go through it. And I started to pick up on a pattern as I was going to theme parks and sailing and doing other things: when people create web apps, great. Even back then, it was not too different than now — you can get to a web app faster, but it’s still kind of the same shape. You can get to the admin, you can get to all the REST API calls, connect platforms, all that. As soon as you try to build AI with AI, it just doesn’t work. And I say “doesn’t” in the present tense, because it didn’t work then, and it kind of kept not working, and it still doesn’t completely work. You get a lot of carbon copies of the initial experience, but back then especially, people were running into brick walls. And I started to see things I really don’t like to see in tech, where people were getting excited — it’s almost like a gambling type thing, where you keep pressing enter and enter, especially people that were not technical, being like, well, now I can do everything.

Jonathan:
Yeah. Like a slot machine.

Nate:
They were taking their life savings in certain cases — like, I saved this up, now I’m gonna pay an engineer to finish what I started. And all this didn’t feel like — wait, this technology, which I do believe in, even back then and before that — okay, it can be magical. It can unlock our creativity and knowledge in ways that are positive. But that didn’t feel positive. That felt extractive — like the companies are benefiting, but the people are either watching them spend money that should not go to that, or it’s time, and this recurring loop that doesn’t seem like a virtuous cycle. It seems more negative. So I just felt that for a bit. I talked with other founder friends about what they were building, and they were hitting similar walls, where building the web app, building the basic AI API call was fine, but actually building a full app, they were just spending a lot more time and money. And I would actually say it’s gotten worse in a way, even though the technology has gotten better — we’re now all in these loops, and they don’t feel great, but we can’t really talk about them as much, because we’re almost waiting for some model to improve everything for us. I’m speaking broadly. But anyway, I was picking up on these patterns, and I’m a pattern guy. I like seeing, hey, there’s this combination of human impact and, you know, anguish in an area where there’s something they want to get to and they can’t — and also just technology, and what’s working or not. So that started me down the direction of, quite frankly, a lot of research into why, and prototyping. And I did that for quite a few months before I really picked up on, let’s build a company or a product. So it wasn’t like an instant thing that RunType came out of.

Luke:
So what does RunType do?

Nate:
So RunType, at a very high level, you could call an agentic development platform. We’re really trying to make it easier to take that vision and that ambition of, I want to build a product — and product definition for me is something that another user is going to get value out of. It’s not just a personal agent. It’s something that another user is trusting you to do a good job with. We want to help the complete lifecycle of development there. On one hand, you can use RunType to basically materialize an AI product, from the back end and front end, that you can test and feel and see if you like it, and put the guardrails in place, pick the right models, know all the costs. You can also hook your existing agents in and help tune it for your users, and basically create — let’s call it a simulation of different architectures — and say, what’s best for my user there? So typically you’re bringing in use cases, you’re bringing in what’s good and bad, what you don’t want the agent to say, and whatnot. So we’re both for net-new and existing deployments there.

Luke:
Could you say it’s a harness?

Nate:
Yeah, I think for product teams it is a super-powered harness, if you will. For engineers, it’s closer to a toolbox that helps you get the most out of your agents. So it really depends on the persona that’s using RunType.

Luke:
So, on vibe coding for a minute. Some people like to use this analogy, saying that vibe coding is to human coding as C code — compiled C code — is to assembly. The analogy being that one time we wrote in assembly and we understood every single little bit of the computer and how it worked, and then C was an abstracted layer on top of this. And now vibe coding — people are complaining that it’s not as good; if I look at the actual code, it’s not how I would have done it, and it’s not as high quality. But the vibe coded stuff is an abstracted layer on top of what we would have done previously, in the same way. Is that a fair analogy? Is that how we should be thinking about it? Or is it still worthwhile, as developers, getting our hands dirty and diving into the code base and really understanding what it does and how it works?

Nate:
Yeah. I mean, one, we’re learning how it affects us. I think anybody that has, like, “this is exactly how it will be in the future” probably hasn’t dove far enough into the clear pain points of developing in pure AI — and also choosing, for a stint of time, to not develop in it. You kind of need to live it for a bit. We’re still all living it. But I said gambling earlier, and the more and more I use it, the more it definitely feels like — sure, you can make a lot of money gambling by luck, if you walk into somewhere and you’re like, oh my god, the first slot machine I went to made all this money. And then you walk out of there and you tell everyone, if you just take in five hundred dollars and you put it on black — you know, that analogy. Sure, for that one person, it worked awesome. For most people, if they’re really good at getting a return on gambling, it almost is a full career for them, where they have to analyze everything. And you get more into skill — like poker, people that are not just going and doing slots. Anyway, that analogy is very close to reality, where the more you get into it, the more it’s about what parts of it are skill that you can tune and get really good at. I think there’s a lot of aspects of vibe coding — which is now turning into a big part of regular coding and architecture and design, even — but you have to put the time in, just like you would put the time into reading and writing code. And now it means that you need to know when to jump in and change course. Don’t one-hundred-percent anything in terms of how you build. Go, hey, this doesn’t feel right — I’m not going to just keep going down that path, I’m going to poke. And if you don’t have any idea on what to poke on, that’s an issue, right? It doesn’t mean you have to be a genius engineer at every part. It does mean, shouldn’t you be curious? Shouldn’t you try to improve that skill? So I think it’s this spectrum, where everybody feels like it’s magic at first, and then as you go, I think we’re gonna have a different development cycle of: this is what I need to really know, this is what I need to have taste on, this is what needs to be unique, and this is what just helps me push the vision faster. And that’s gonna grow over time, to where we’re gonna be able to trust more and more components of it. So that’s generally how I think about it. I have my own personal rant I could go on about how I think it feels day to day — there’s a lot of frustrations there, where there’s an abundance of opportunity for us to make these things better, not just be like, okay, it’s painful, we’re gonna let it happen.

Luke:
Friend of the show Tammie Lister likes to call it swear coding. There was a time when I was doing a lot of vibe coding, Nate, and I stopped, actually. I’ve stopped vibe coding. I don’t do it anymore. And the reason why is because I love programming. And one day I was building an app in Swift, and I forgot how to concatenate a string. I just couldn’t figure it out. I was like, I used to do this constantly — what’s happened to me? And there is this documented skill atrophy issue. How should we be thinking about that? It seems the more we vibe code — it’s not like riding a bike, right? We lose our skills. We lose our ability. We lose our memory for it.

Nate:
I mean, I think it’s a slightly different question here: what do we reward? And this is more of a broader question. I think right now we reward this instant gratification, combined with the ability to very effectively market. Right? A generic example would be, every time a model releases, what do you get most of on X? You get people building full-blown games, right? They don’t understand any of that, but they are getting some value out of playing it. They’re gonna build another game, and all of that, right after. They could get a ton of views, they could get people reaching out to help them — even, maybe, help me build my game. But it’s almost talking into the same ecosystem of other people building quickly. They’re getting rewarded — but is that a thing to be rewarded long-term? And I’m not talking down about that — I think it’s fun to build that stuff too. But I know what I’m rewarding personally. I’m taking my time and doing almost like a pleasureful ideation experience. I’m not gonna go sell that game. I’m not gonna go take on someone that put the effort in to actually design a game, right? So that’s what I think about when it’s like, should we vibe code or not? I think it has a place. I consider it more AI-assisted coding, and how autonomous you are on the spectrum — I don’t call it vibe coding or not. It’s really the use case, it’s what value you’re trying to get out of it, it’s how production-grade it should be. And if we’re only talking about vibe coding, it’s more personal to me — when it’s, I just want to see what I can do, that’s fine. When it becomes multiplayer, and something somebody else depends on — they’re running their life on it, and let’s say you get to the point where it’s a robot in your house — that’s a different type of thing, right? At some point, even if you automate the engineering of it, it’s different.

Luke:
All right. But Nate, what about the cognitive atrophy, though? What about that skill and mental thinking atrophy — we’re not as good as we were before, the more that we rely on AI to think and do and program for us. Should we be worried about that? Should we be concerned about it, or should we think of it as a trade-off?

Nate:
Well, again — what do we reward? Let’s use a different thing. Look at previous eras, when new interfaces came out and new ways of programming — we’ve been through that a couple times. And I think it always is uncomfortable to be like, what’s taken away from me? What do I not need to understand anymore? It takes time to process that. I would think, if you look through each ebb and flow, you were rewarded to understand, what is the new interface? What does it need? Does it need a new programming language? Does it need me to be down deep into how the hardware works? That’s more of what I mean by that. So my answer is, I can’t broadly say that. What skills, depending on the problem you’re working on, differ. For website design — web dev — there are parts of it where it’s probably okay for it to atrophy, if you have taste and if you have QA, and the actual coding of it is less important than the output being unique and real for the person you’re developing for. If your taste atrophies along with your skill to understand what you’re doing — well, what are you really building? I don’t know if that makes sense, but that’s kind of how my brain’s currently working through that.

Jonathan:
It’s been interesting — I’ve spent some more time with you over the past couple of months, Nate, and got some quality chicken time as well. And in thinking through — you’ve made a big move from Puerto Rico to San Francisco, and you’re in it there in Silicon Valley, and all the things that are happening. There’s a lot of hype to all of this. And part of what’s impressed me in observing and listening and watching what you’ve been building, and why — so, none of us are old, so to speak, but we’re getting older, and there’s this generational thing where we’re watching the next generations come into this stuff who don’t have the background in building things. What I’ve appreciated about what I’ve heard in your experience so far, Nate, has been this focus on still putting in the effort and doing the hard work. You’re not trying to just do the slot machine. But as I’m listening — it’s not that the slot machine is inherently a bad thing. It’s: are you using the slot machine to build something that other people are going to rely on? To me, that’s at the heart of it. I don’t want a robot in my home that was built by someone who just slot-machined their way to it. That’s not gonna work. And I think that is what’s happening in the product world. There’s people who are springing these things, and they seem fully formed, but then once you get into it, the substance isn’t really there. Things break down, and they’re like, well, I’ll just patch it. And I really enjoy playing games, especially games built by these indie developers. And it’s been interesting for me to watch those who are vibe coding versus those who aren’t. I’ve seen some things where I’m like, okay, there’s some creativity and effort here, and I love how vibe coding has enabled them to get something to market that they couldn’t otherwise. I’ve also been pretty unimpressed by a lot of it, where I’m like, you’re just stacking problems on top here. You’re responding quickly to user feedback, and trying to iterate and stay responsive, but you’re actually not doing product work here. There’s a lack of maturity on their part, where it’s like, you don’t really know what you’re doing, and these tools are actually causing more harm for some of these indie developers, where they’re like, well, I can do the thing the user asked for — but what they’re not taking into account — there’s a taste component that’s lacking in that process.

Nate:
Well, as part of that, though — this is what’s interesting. How do you generate taste? How do you grow taste within yourself? And I think experimentation, repetition, feedback is part of it. That is where vibe coding, if you use it correctly — I’m talking especially for someone that’s straight out of college, they’re just moving quickly, whatnot — sure, it’s great. It’s just knowing the limit of when do you graduate. Not meaning you never have AI do something, but you start to learn: hey, this is where I need to actually get an opinion that’s different than the average. Because that’s what we’re doing by default — we’re averaging out all of the great human knowledge into something. I don’t frown on that; I just know it’s an average most of the time. How do I not make it average? I have to make it mine. Right? And that goes with anything. Let’s talk about a design library. If you start up and you get shadcn as the default for every web app — does that mean shadcn is bad? No. Should you understand when you should make it unique? Yes. Even the creator has said, I built this library for people to make their own, as a foundation. But people don’t always do that. People just one-to-one it. So it’s a human trait, I think, to take something great at foundational tech and be like, I want to make it mine, but I’m not really gonna change anything, and I’m gonna try to sell it or do whatever. And again, I think it’s a spectrum there.

Jonathan:
This is something I continue to struggle with on the writing side. You guys are talking about coding, but a lot of people are using AI to output words for them. And the heart of my struggle is, when I’m reading things that are good — like, it’s good, it’s technically pretty sound — but I’m like, this isn’t that person’s. I come to a line, or I come to something in there, and I’m like, that’s not you. Maybe you even read it, but that’s not you. So I find myself — I don’t use AI anymore, and barely did, for my own writing. There are things where I’ll use it — I’m still trying to figure these things out. But the heart of what I notice when I see things from others is, yeah, there’s a quality there, but it’s not you. There’s this lack of ownership. And to me, there is a baseline where it’s like, you have to have at least read it. But what I’m noticing when I have things generated and I go through it — there’s almost this confusion in my head where I’m like, okay, those words were good, but what is me? Where does my taste come into this? And I often find that it’s easier to just throw it all aside and start from scratch and think through, okay, what are the words that I wanna use? But I’ve also invested a lot of time and energy and pain into thinking about words over the years, and a lot of people won’t.

Luke:
Filling a blank page is a skill, you know — being able to stare down a blank document or page and get started. That is a skill worth having, I think, in and of itself. And I was bringing up skill atrophy before — it’s the same sort of thing you find in the workforce. A lot of these young kids, these Gen Z-ers, are really struggling with things like writing emails and communicating in a written way. I find that a little concerning.

Jonathan:
Well, and just to add to that — part of what I’ve seen recently is that the folks who were struggling with poor communication suddenly seem to have —

Luke:
We’re totally old, by the way. We definitely are old.

Jonathan:
I know — as I’m hearing that, like, we’re getting old. I’ve seen folks go from struggling with basic communication to suddenly having, quote-unquote, excellent communication. But it’s not them. You could be using it as a tool, but then you’ll see this disconnect from how they write to when you actually have a conversation with them — it’s very different. So I’m like, that’s concerning to me. It goes back to that point of ownership. I had this vision the other day of, man, it feels like there’s a not-too-distant future where — you know, having a model in your head, whether you’re listening, or people going on a date or something and asking AI to tell them what to say. It seems so wild, but I can see another generation coming up where that just seems normal to them. And that is concerning to me. And at the same time I wonder, is this just me getting old?

Luke:
And also, it might be like a San Francisco effect. The kids I see — more Gen Alpha and Gen Z, the kids that I interact with — they actually really reject AI for a lot of things. Have you noticed the same thing, Nate? Or is it different in San Francisco?

Nate:
Yeah, actually my brain went to region and what I see. Look, it’s not a one-to-one ratio, but a lot of times, what happens in so-called Silicon Valley with technology is a trend that takes a little bit more time adoption-wise — there’s a creation of technology separated from the widespread adoption of it, and the behavior and the trends. It does seem like it’s starting to slowly bleed out of SF. It’s kind of a combination of what seems like internet culture and social media — like, by default, you’re always on there — and then with AI being the aid there, to rapidly broadcast your thoughts and ideas and whatnot. And people that get into that — it’s hard to get them out. Because it’s not all negative. Some people have trouble connecting because they don’t know what to say, don’t know what to create. So it makes it easier for them. But likewise, it kind of brings down the average, maybe, sometimes, of people where it would have been easier for them to stand out, ‘cause they were the ones that were out there. So yeah, I think that it’s a little bit inevitable — but what’s not is the shape of it, and the shape of the things to come. And that’s what I think a lot of what we should be focused on is. Hey, okay, if we’re uncomfortable — like, Jonathan, you’re saying, I don’t know if I want that to happen — well, how can you shape that? We can shape it. It can happen to us — we can let the devices, and what’s built into us as humans, drive behavior that makes us uncomfortable — or we go, hey, let’s build technology, policies — let’s build just the way we want the world to work.

Luke:
I’m sorry, Nate — super old man take — and say, we don’t have a lot of say over it, actually. And I know this is a sort of jaded opinion, but technology companies, big tech — it’s all about hooking users. It’s all about, I guess, commercialized addiction. I find it really hard to blame people for being addicted to their smartphone, because it’s the big technology companies, especially the social media companies, that I would hold accountable for that. That’s what they’ve attempted to do, and they’ve been really successful at it. And so when it comes to, we can decide how we use technology and how we shape it, and it’s up to us — nah, I don’t think so. I think it’s sort of being dictated by the people who make the software that we become addicted to, a little bit. Sorry, that’s super jaded.

Nate:
I knew you’d say something like that, Luke, by this point. I checked that off on my little bingo board, yeah.

Luke:
You did. I love being predictable. This is, I think, the all-time classic old man thing that’s been around since the ancient, ancient times, and I think it’s just as applicable here — which is: maybe it’s important that we do hard things. Maybe those people who are really struggling, and AI makes it easy for them — maybe that’s not good for them.

Nate:
I mean, I would say, to flip it around on you: the hard thing is to prevent a handful of companies from controlling our being. That’s one of the hardest things. You just said it almost seems like you can’t do anything. Well, if you think we should do the hard things, I’m agreeing with you. I’m saying we should be in it. We should help shape the world there. And the benefit I think we have — anybody that’s been even, let’s say, just a decade in tech, you have ten years more experience than someone that’s coming out and AI is their whole world right now. It’s on you to say, okay, how do we shape it and control it? And the way I see it a lot is in very simple terms: you want to know what the different models are capable of, and if you need the most expensive version or not. You want to know who has your data, and what you’re comfortable with or not. You want to know, can I build my own representation of what a product is, and get utility out of an assistant or whatnot, using more of the open technologies, and blending it? Because, look, we have our data in commercial products like Gmail and all that, but you can make the choice every step of the way to understand, hey, I made the decision here of what’s happening. And eventually — and this is something I’m really passionate about, and there’s some bias here, if I want to help people build the products of the future — we need to get comfortable with supporting not everything in the box of one company. Because a lot of people I see that are against where the world’s headed, with a couple companies owning it and AI — they’re also using Claude and OpenAI predominantly, if not the only thing they’ve ever used in AI. And then when you poke a little bit — well, would you use somebody creating something that wasn’t Meta or OpenAI or Anthropic that supplied the same use case? And a lot of times it’s like, I don’t know if I’m comfortable with that. Well, okay — so you’re not comfortable with the big companies running away with everything like they did in the previous eras, and you just kind of succumb to the momentum. But you have a choice here to use other creators, other people that are taking that and shaping it a little bit. It doesn’t mean there’s zero data retention — a lot in the cloud and internet, it has to go somewhere. But it does mean it’s not a hundred percent. So what if you could give only fifty percent of your data away, and you can support somebody that created their own instance? And a lot of this goes back, to me, to WordPress and all the open communities. Hey, try to get people controlling a little bit, right? When you go and turn on WordPress AI and you want to control your own experience — well, try an open model. You know, little decisions there.

Luke:
Yeah — that distributed nature of the web, and the idea of a distributed AI ecosystem. That’s a sentiment I can really get behind.

Jonathan:
Hmm. Man, this has been a lot of fun. Nate, we’ve had multiple conversations over the years, and it’s gonna be interesting to watch what you do with RunType. It’s also been fun to see the investments that you’ve continued to make in open source, with the persona.js project that you guys have been working on. And we don’t know where this stuff is gonna go, right? Having a conversation like this again in a year could look quite different. I am seeing this clearer vision, Luke, of us getting just older — and I don’t really see you changing too much.

Luke:
Yeah. It’s going to stop being “get off my lawn” and it’s going to start being “don’t send me AI emails.”

Jonathan:
Right. But there’s — I don’t know, there’s something in all this, and Nate, I’m appreciating the background that you’re bringing into this. I think of it a bit like there’s a gold rush going on right now, right? And I think there is gold to be found — there’s a lot of hype through all this as well. From just a business perspective, I think it’s better to be in the making-pickaxes business, making the tools for the gold rush. But what I’ve appreciated about this, in knowing you over the years, is it’s even more than that. There’s more to it philosophically. It’s not just building the tools — it’s also thinking about how people are using this stuff, and what we’re rewarding. And I like that sense of glass-half-full I’m hearing in you: that with some of this open web stuff, there is still opportunity to guide. It’s not all just set in stone.

Nate:
Yeah — like another framing, and this is for Luke: a frame would be, this is like the GeoCities era, if we make it that. Let’s get more people creating. Let’s get more people owning it. And yes, it will be shaped — but it’s not GeoCities, if we’re honest. It’s more like we skipped past the era of everybody experimenting. It went from ML and data scientists and a lot of people in tech to over-applying it to everyone, instead of just having fun and learning and whatnot. Vibe coding’s a bit of that, but it has so much business in it, by nature of what I can hype. So for me, I get excited about bringing an era to form where it’s, hey, I can do that, right? And I can learn from it and apply it, because all of us learn from just hacking — just having fun there. It’s a different form of hacking.

Jonathan:
Nate, what’s the best way for someone to get in touch with you?

Nate:
becomevocal on X — which I always try to get back into and I’m not perfect at, which is funny, because my tag is “become vocal.” But I would say there, or Nate at RunType. Even if you just want to reach out about a product you’re building and it’s not AI — I just like to talk to builders, so happy to talk to anyone. Also — I wanted to get it out today, but I didn’t — we’re gonna have a release candidate up on GitHub of Persona for WordPress. So it’s gonna be open source. Out of the box, it will basically make WordPress AI easier to render and build your own assistant. Again, you don’t have to use RunType — if you do wanna use RunType, it supports all the open models, all of that. So if you want to experiment in either direction, and contribute and give feedback, I would be happy for anybody in the audience to do that. I’ll hopefully be sharing more of that soon.