394: Taylor Otwell — The (Quite Entrepreneurial) Creator of Laravel
Download MP3Hey. It's Arvid, and this is the Bootstrap Founder podcast. Today, I'm talking to Taylor Ottwal, the creator of the Laravel framework for PHP and an outstanding entrepreneur who figured out how to turn open source software into Lamborghinis. Seriously, Taylor's approach to building not just a framework, but a whole ecosystem of businesses on top of that framework is one of the most impressive open source monetization stories I've ever seen. It's quite fitting that Laravel has therefore also embraced the sponsor of this episode paddle.com.
Arvid:Paddle is a merchant of record payment provider, and they are deeply integrated into both the open source library Cashier, which is free, and the four pay payment portal Laravel Spark, one of the many ways that Taylor is monetizing his work. So it's open source and for pay. It's a very interesting story. So if you want to monetize your own software work without having to deal with stuff like sales tax and expired credit cards, give Paddle a shot at paddle.com. They're supremely good at making your founder life easier so you can build your business on solid foundations.
Arvid:So let's talk foundational frameworks. Here's Taylor Ottwall, the creator of Laravel. Congrats, Taylor, on half a million deployments on Laravel Cloud. That is amazing. How how does it feel to see people building on your own platform at that scale already, a couple months after you released the thing?
Taylor:Yeah. Wow. It's pretty surreal. And it's especially surreal thinking back to like how it all started, which is basically like a casual conversation in Amsterdam by a few of us at There was not many of us at Laravel at the time. There was probably only nine or 10 of us.
Taylor:And now there's like 70. So just thinking back to that conversation where Laravel cloud was just an idea and we hoped we could build it, you know, we hoped we could figure it out and crack it to now where it's live and, you know, thousands of people are using it and there's, you know, so many people deploying on it. Really it's pretty mind blowing, but I'm super happy with where we're at. And honestly, I think we're just kind of getting started with it. So I'm even more excited about where we're heading with it.
Arvid:Yeah. I bet. That is such an accomplishment, I think. Like, just seeing you develop both the idea and then teasing it a little bit at the Lericons and all of that, it it felt like a it was a concerted effort, but it also felt very natural in in the way that it was built. That was really cool.
Arvid:And looking at the Laravel ecosystem in particular, like the products that existed prior to cloud, I can't help but feel like you worked your way, you iterated your way to cloud, to forge and and Voya way, the the hosting and deployment, did did this make it easier to actually build a full cloud system? Mhmm. Mhmm.
Taylor:Yeah. That is kind of interesting. Like, Laravel cloud and fully managed Laravel deployment and hosting is something I've danced around for like the last ten years it feels like. And when I built Forge in 2014, I really would have built it more like cloud as like a fully managed platform if I was able to. It was just me, you know, like hacking on this idea in my free time and I didn't want the risk of managing all of this infrastructure as like a solo bootstrapped founder.
Taylor:So, I built Laravel Forward for people could bring their own cloud account. Hey, if there's something happens with the servers that they're running the apps on, that sort of like DigitalOcean's problem, you
Arvid:know what I Yeah, sure.
Taylor:Or AWS's problem. But yeah, we have been kind of working our way towards this with things like Forge and Envoy and Vapor for a long time. And you know, building something like Cloud just takes a much bigger team and much bigger, risk honestly. You know, you're just taking on, you're biting off a lot more, when you take on a problem like this. So we finally got here, you know, in 2025, all these years later, and, you know, it's an exciting time.
Arvid:It it sounds like you unlocked something inside your founder brain for this. Right? Like, did that happen? Like, what what did you do? Because honestly, I would be so scared lead a team of 70 people, let alone seven people.
Arvid:Like how did you massage that into your mind? I mean, as you
Taylor:may know, I bootstrapped Laravel from 2011 until 2024. And in that time I launched, you know, five commercial products, Forge, Envoy, Vapor, NovaSpark, and something like 27 or 28 open source packages in addition to the framework itself, which is like the core of the whole thing. And in 2024, or really I guess 2023, I felt like I was sort of like at this crossroads with Laravel of like, man, I've really built almost everything I wanted to build, every idea I had. The whole idea for Laravel was only to create a tool to help me build startups faster. And then that became like the meta business that ended up becoming my career.
Taylor:And so it was almost like, I can either coast and be like, okay, that was a solid run, know, like an open source few people are able to enjoy that level of open source success. Or there's like sort of a next level that we could go to, which is building something much more ambitious, also much more risky, but maybe a much higher reward with things like Laravel Cloud and Laravel Nightwatch, which is coming out soon as well. And you know, that was just kind of what I wrestled with for a long time and obviously decided to sort of swing for the fences with Laravel Cloud and try building a business with investment money, you know, to build these much bigger, more ambitious things that I couldn't build on my own without great personal risk and liability. So, and I basically, I felt like I owed it to the community is what it boiled down to. I felt like if I just sort of coasted, it was like, okay guys, that's sort of the peak, know, that's where it sort of tops out.
Taylor:I felt like that would be just such a bummer for like the tens or hundreds of thousands of Laravel developers in the world that have been along for the whole thing. And so that's ultimately what led to deciding to pursue cloud.
Arvid:Interesting. Yeah. There's this this concept of the benevolent dictator for life, right, that always exists in the open source world. That can go either way. Right?
Arvid:Either you can just do whatever you want as a dictator or you can try to be the benevolent part and serve your community. That's an interesting one. I do wonder, how do you balance this? Because particularly with cloud being a very much for profit enterprise and having investment, having funding, and on the other side, have this open source community of people who often have the expectation that things should be free and things should be designed by open source contributor committee. How do you balance expectations between these kind of groups of people?
Taylor:We really just try to let, I think, our behavior do the talking. You know, I think that's really the best way to do it. So like we have full time staff working on open source and we release open source every week and honestly at greater frequency than we did before. Just because we have more resources behind it. I think Laravel is also in this really unique position with cloud and forge to where we almost have like both sides of the coin, you know, like we have this fully managed platform for people that just want to kind of deploy and they never want to think about servers.
Taylor:And if they pay a little bit of a premium for that at the end of the day, that's fine. And then we also have Forge, which is like this kind of semi managed server management platform, which we're rebuilding entirely for LariCon this summer, which is gonna be cool. And people can bring their own cloud provider or whatever they want and you know, throw 50 sites on a single server and really like get a lot of juice out of that server for the amount of spend that they're putting into it. And you know, I think Laravel is unique to like offer both, like whatever's better for you, we're just trying to help you deploy your application and enjoy using this framework and you can kind of take either path. So open source like we know and I know and I think the company knows, if we don't have Laravel and we're not committed to open source in the community, like no one's gonna be using this stuff.
Taylor:So we have to be more committed to that than ever, basically.
Arvid:Yeah. I've always admired, like, the strategy that Jorg just laid out. Right? You build products that help people use the tool that you also build, and then you use the tool to build the products. Like, there's a lot of synergistic effects between these.
Arvid:That makes sense. And that to me is something that is, well, not completely unique, like other open source communities try to emulate this, but it is very efficient in the Laravel world. You just mentioned five products. I use four out of these for my software stack. So you said you built five products.
Arvid:You had a five for five success with these products. Were there any along the way that you tried to build and they didn't work? Or is it all VaporForge and Voyeur and the other things? Did that all work like from the beginning? Like, that that is such a such a rare thing in the serial founder world for that to actually happen.
Taylor:Yeah. It is pretty crazy. There was one product that I built all the way through that I never actually launched. So I don't know if it would have succeeded or failed. And it was actually called Laravel Cloud at the time, but it was more like a Laravel Forge reimagining that helped you deploy like load balanced VPSs and you had like a YAML config file.
Taylor:It was just a little bit different, but I got it all the way to completion at a UI built by Steve Shoger, I believe, from Tailwind because he worked for Laravel at the time. And at the end of the day, I was like, this just isn't different enough from Forged to actually be compelling. It's actually like sort of confusing. There's no clear difference between the two products besides a few like minor things. And I just like didn't even launch it.
Taylor:I just tabled it, which is crazy because the whole thing was built, the back end, the front end, everything. But you know, just think it would have kind of confused things. So other than that, you know, everything kind of worked. But the whole way though, was just solving my own problems. I wasn't really trying to like, in many ways I wasn't even trying to build products.
Taylor:I was just like, especially with something like Vapor, I was just personally interested in AWS Lambda and could I get Laravel running on it? And could I like build a deployment platform on it? And I just, I went down that road for like six months or whatever, and I was like, well, I guess I'll ship this thing. And you know, it just kind of worked out. So a lot a lot of kinda happy accidents, some pre planning and masterful thinking, but a lot of happy accidents as well.
Arvid:Yeah. That that sounds like coding to me. Yeah. That's for sure. And and software entrepreneurship for that matter.
Arvid:Right? Like trying to figure things out as you go. I I do wonder now that you have cloud, now that you have these massive numbers of deployments, and probably also other kinds of products that people are trying to deploy than you may have seen before on Forge or whatever, How does that impact the actual framework? Like, how does your experience with this at scale impact your plans for the framework in the future?
Taylor:Yeah. There's been a few things. I wouldn't say they're major things, but like when we start thinking about the scaling systems in cloud, we need maybe more insights into what's happening in the queue. For example, like how long are jobs sitting on the queue? How long are they running on the queue?
Taylor:What's the oldest job on the queue? Like that kind of info that we haven't exposed before, maybe unless you're using something like Laravel Horizon, but for all queue drivers, we need that information to like intelligently auto scale. There was a few other minor things we added to the framework to make like, you know, storage easier with cloud or something like that. But honestly, didn't change any major pieces of the framework. I think what's interesting for me on cloud though, that I didn't expect in terms of like the types of applications we're seeing or maybe customers is, when we built cloud, I kind of wondered like, or at least one of the fears I had was everyone's already kind of deployed their Laravel apps on Forge, know, and so why come to cloud?
Taylor:But I think what we're discovering is there actually was this like contingent of a little bit larger applications that maybe were rolling their own Kubernetes at some company, and they're having headaches with it all the time because it's complicated. And they would love to run something fully managed, but they felt like Forge was not quite enough for them, and cloud kind of hits that sweet spot. So we are seeing customers like that, that I didn't expect, which is really cool to hear their stories and to talk to them and to try to get them on cloud and help them be successful.
Arvid:Yeah. That that sounds like you're you're getting into a new segment of customers here.
Taylor:Yeah. Right. A little bit.
Arvid:That's that's really fun and probably, like, highly promising for the success of the platform.
Taylor:Yeah. Yeah. I'm excited about it. It's fun.
Arvid:Has people using Laravel on the cloud surfaced something about Laravel that you didn't expect to see? Like, is there is there any surprise in there, or has there been?
Taylor:Oh, gosh. You know, we see all sorts of things both on cloud and Forge, know. See things
Arvid:Don't look at mine, please.
Taylor:We see things like the the typical stuff you might expect with a web framework, like a a web request that's firing off 500 database queries and the page is slow and depending on what database or cache setup they're using, that adds up to a lot of latency or whatever. So we see things like that, maybe more so than we expected, honestly. So, you know, there's a temptation to want to be like, okay, well you should fix your application, that's not our problem, but that's gonna continue to happen. We have to make it fast even if the app is not ideal, you know? So that's one thing we're focusing on right now is like, how do we make it as fast as possible even if the app's running 400, 500 database queries in a single request.
Arvid:That's that's interesting. Yeah. I I guess that is also a a problem that you will never be able to get rid of. Right? You always have beginners.
Arvid:You always will have, like, AI written code that is super weird and pretty and non performant that that will be deployed probably automatically because there's some weird auto PR merger or whatever. Interesting. I guess that you have to solve on the hardware level or on the the optimization level. But I wonder because now with with Nightwatch in particular, you get more introspection introspection into, like, how this works. So is this gonna be something that you're gonna try I don't wanna say to push, but to suggest heavily to users to to kind of use by default just to see their performance and be able to optimize it?
Taylor:Yeah. I I think you should. We've discovered so much honestly about our own apps. We've been running Nightwatch on Forge. And to give context, Nightwatch is like a application observability tool for Laravel that shows you all your performance metrics for request queries, all that.
Taylor:We've been running on Forge and we immediately saw things we didn't know about. Whether it was like these outlier requests that are super long for certain customers, and many times those are your most important customers because they have the most servers and data and all of that. Or pages that were doing more queries than we expected. We actually had a situation the other day on cloud where we were having trouble debugging a problem, I believe, and so we actually put them on night watch early to help figure this out and we're able to like immediately see what was going on, which was super cool. Nightwatch should be coming out in early June.
Taylor:It is going to have a free plan actually, a totally free plan where you can kind of kick the tires and see the entire product without really any feature gating or anything. So everyone will be able to kinda get a taste of it and see, like, if it works for them or if it's helpful. I think it's it's pretty crazy helpful in my opinion. So I'm excited to get it out there.
Arvid:Yeah. Like and and some some kind of observability tool, APM, whatever you wanna call it, is, for me at least, something that I try to get in as soon as possible just to avoid these kind of scale issues because it tends to be a scaling issue. Right? Like, things are slow. That's fine in the beginning, but they get slow in aggregate.
Arvid:Well, that's a problem. Right? So that's that's really cool to see And also been something that I've been personally having trouble with in the past. Like, just even to get Exentry or whatever integrated into a p PHP application well, that's hard. Like, there's a lot of work that needs to be done that has very little to do with the actual application and a lot to do with the server management side of things.
Arvid:That's what you guys probably do much better than I. Right? So for that reason, I'm looking forward to it. I can't wait to get my invite. That's tough one.
Arvid:Because I'm really looking forward to not having to deal with this and also knowing that the people who deal with this actually know what they're doing. That to me is that's the great thing about the Laravel ecosystem to begin with. Like, you you mentioned Spark earlier. I I process all my my payments through Spark, right, through the the billing system. And Nova is is also my back end that I use for administrative stuff because that is just the best to integrate into a Laravel application.
Arvid:I think you've doing a really good job at finding the next thing that people use something else for right now because there isn't anything, but actually want something done by you. Do you think this is going to change with cloud in terms of that you're not going to have standalone products anymore and it's all going to be part of cloud? Because that's, I think, a fear that a lot of open source people might have that it's all going to be bundled into this and then not be able to extract it into a standalone deployment anymore? Like, what's the strategy for you there? For future products, I mean.
Taylor:I don't think we would try to force anything like that. I think we'll honestly have our hands full with just the core feature set of cloud for a long time. So to be honest, we haven't thought a lot about like future products that might be standalone or built into cloud. It's just sort of too far out there on the timeline. But even Nightwatch is a good example of like, we could have sort of built the observability stuff into cloud and said, you want all of this good observability stuff, need to run on Laravel cloud because that's most commercially advantageous to us.
Taylor:But we didn't do that. Obviously, it's it's a whole separate product that you can use whether you're on Forge or Cloud or even just rolling your own AWS deployment. You just run the Nightwatch agent wherever you are and we start pulling in the data. So you know, I've already been conscious of that and it's something I want to avoid is sort of like, I think it's very obvious when companies like artificially force you into certain things that are only that way so that they can get the maximum amount of money from you. Instead of like what's actually good for the user or the most users or what's gonna be good for us long term.
Taylor:Yeah. I think we'll have our hands full with these two new products for a while anyway.
Arvid:Oh, for sure. And I think they are probably, yeah, the most ambitious and most like nuanced products that you've built, like, ever ever since you started with all of this. I I love this. Like, when I when I heard the the Nightwatch announcement at Lericon, like, I was watching it through the stream because, you know, it's busy at another conference. But I I love just the idea of you finally tackling this because it always has been something, I want this.
Arvid:I hope they're gonna make it. And and then it's just gonna naturally happen because you you also have the the other system that needs the observability to be performant.
Taylor:Right. Yeah. Me too. Like even even just like one screen, if like for me, like the log screen of Nightwatch, anytime I'm starting up a new Laravel project, I would always just almost kick the can down the road of setting up proper logging because I knew it was just gonna be a pain and I didn't wanna mess with it. But just even if I only had logs in Nightwatch, I'd be happy with that.
Taylor:Like that's just so much better and a level up of the experience for me.
Arvid:Yeah. You should see what I'm using. I I use, like, an SSH connection that has a tail on the Laravel logs file in the storage directory. That's my logging. Like, that's running right here.
Arvid:Like, it's got terminal somewhere.
Taylor:That's nicer looking than a lot of solutions I've seen.
Arvid:Right. But that's not how it should be. Right? There there should be, like, from the start, a way for you to opt into something that that works for you, has data retention and searchability, that kind of stuff, that that matters a lot to me. And I'm I'm glad you're building this.
Arvid:I do wonder, and this is probably something that is a very recent thing, how much AI in all kinds of ways Laravel will have in the future. Right now, AI and Laravel to me is really just I ask my ChatGPT and my Clawds to build me Laravel stuff, and it is good enough to build this and understand, like, what Eloquent is, what the Laravel, like, directory structure and and what jobs our queues are in that. Like, AI can build Laravel apps pretty well, but I wonder what your perspective is on both using AI to code with and AI as a part of the larger level ecosystem. What are the thoughts there at the moment?
Taylor:Yeah. So I'll start with, like, my view on AI coding in general, which I'm really bullish on. I think that probably all programmers will use AI assistance in the future. I think that's pretty inevitable at this point and it feels like we're almost already there basically. And I think it's great, honestly.
Taylor:Like I think it's awesome. I wish I had those kinds of tools when I was learning to program. I think it's awesome for writing the language you know, but I think it's also awesome for learning new languages and spiking out something in something you're not familiar with. So, very bullish on that and think that will continue to get better and more pervasive in the future. As far as AI and Laravel, I mean, I think about it in a couple different ways.
Taylor:One thing we're kind of kicking around now is, I think you're right that like the current tools actually write Laravel pretty adequately and it's totally pretty much fine every time I use it. This kind of stuff is changing every day. Like even just yesterday, I saw like the Cloud Code SDK and the new GitHub Action stuff. So it's just like a constantly changing thing. I think I'm interested in like seeing how we can teach the AI sort of like that final level of polish that, you know, you might put on a Laravel app.
Taylor:Laravel has been around a long time, which is great for the AI stuff because there's lots of stuff out there for it to be trained on and code and all of that material. But like how can we get it to write Laravel, how maybe we would write Laravel as a company or be a little bit more opinionated with the latest sort of like best practices and how we would write things cleanly. So we're interested in that. I'm also interested from like a tooling perspective. I think it's going to be really important for Laravel to be the easiest full stack framework to integrate AI with, whether that's like an AI SDK.
Taylor:So in PHP, the most popular one right now is probably Prism PHP, which is like a back end agnostic AI SDK, very similar to Vercel's AI SDK where you can use OpenAI or Anthropic or all these other providers. So with that, another thing we're kind of working on right now is the fastest way to build MCP servers with Laravel. So many people are writing those these days and we'll see if that continues. It's hard to hard to say, because things change so fast, but that's something we're actively working on right now trying to get to one point o is basically an MCP server framework for Laravel that feels very easy to get started with and quick to spin up these ideas and deploy them as well. So I mean, overall, yeah, I think there's going to be more AI related stuff in dev and Laravel.
Taylor:I think it's overall a good thing. I think it's really cool and really helpful. And I want Laravel to be the best way and fastest way and most productive way to build AI powered apps. And I think Laravel has always been an opinionated full stack framework and it makes sense for us to have an opinion and a clean way to integrate this stuff into your app. So I think we'd be kinda like burying our head in the sand if we didn't do that.
Arvid:Yeah. It would be, like, missing the the just the the point of conversation that is happening right now in the conversation too. Like, everybody's trying to figure it out. Like people have opinions. Obviously, they always have opinions and and, you know, in in all different kinds of ways.
Arvid:But, yeah, I like what you're saying because you're not saying we need AI in our product because so many people do. You say we we should facilitate people using AI to build products. That is a very different thing, and I love it. I love the fact that it's not about, you know, like hyping or using the the hype train, but actually building reasonably useful things with a new technology. I I that that is most appreciated, and I agree.
Arvid:Like, Making it easy to build things with AI and through AI, that should be the angle for a framework right now. My big dream is that a framework and Laravel should could maybe the ships the Laravel installer ships with a really, really small language model that is purposely and intentionally trained just on writing extremely good Laravel code. And you can run it locally. It runs in the IDE. I mean, I look at my my PHP storm, and they have some weird internal little code completion things now.
Arvid:I don't know what it is, but I assume it's it's a really small language model. If that would be something a framework can provide and then all frameworks could provide something like this for their code bases, that would be so cool and so helpful.
Taylor:I think we're, like, not far from that, honestly. You know, with, like this isn't local, but with things like the Cloud Code SDK that I that I was talking about that I saw yesterday, the ability to interact with it non interactively like they announced and feed it prompts and context and things gets us close to having like these artisan commands or local things that I actually know how to write good Laravel code, which is I think super cool.
Arvid:Me too. Like, obviously. Like, I yesterday was a moment for me too where I used Junie. I think that's the internal agent that PHP Storm offers now, and and I use it for the time, and it was bizarre. Like, I'm I'm not much into agentic coding because I still want to do my own coding, which is probably a traditionalist perspective at this point.
Arvid:But I gave it a task. I wrote a spec. It took twenty minutes to write the spec. And then I just walked upstairs to get a coffee. I came back down and the thing was implemented.
Arvid:And I was like, this is crazy in the best and worst way at the same time because, like, now the thing is doing my job. All I'm doing is instruct. Interesting, to say the least. Like, how how much AI coding do you use, like, for developing YARL features at this point? Like, how how how does that work for you and the team, I guess?
Arvid:So a lot of
Taylor:the work I do on Laravel is reviewing PRs. And a lot of the changes I have to make are relatively minor and don't require assistance, but some of them end up being pretty big. And what I'm really interested in and what I thought about yesterday when I read about this new GitHub Actions integration with Claude and stuff is like, a lot of times if I get a PR that's sort of non trivial and I have some feedback to give to the author, and then I just kind of have to sit around and wait for that to be implemented. If it ever is, right? Like if they come back, if we're able to unlock that in like PRs as a open source maintainer, if I could just at the agent, hey, this PR is 90% good, here's the things I don't like and kind of want to change about it, can you implement that?
Taylor:And I can go downstairs and get a snack and come back and it's kind of done. That's like insane, you know, unlock. So yeah, I'm curious curious to see how that develops.
Arvid:Yeah. I I see the same thing for security testing or even for just trying to do background improvements. Like, when you were just saying that the thing about how people on cloud built these weird apps that have 500 database queries, Well, what if you could spin up a complete clone of the repo, like have an agent try to fix it and suggest it to them automatically because you're linked to that GitHub account anyway? Right? That to me is smart AI usage inside a product.
Taylor:Yeah. And I think we're super interested in that for Nightwatch. And so some of the ways we're using AI on Nightwatch right now are are pretty basic, like summarizing exceptions or issues or things like that. But you could imagine where we could do something like what you're saying. We know pretty much exactly why the request was slow, either which middle where it was, which query it is, we can see if there was an index on the query, all of that stuff.
Taylor:If we're integrated with cloud, then we can just, we know what repo is, we can send you a PR, we can deploy it for you on cloud. So it's kind of the whole story could be covered, which I'm very interested in exploring.
Arvid:Yeah. For sure. Yeah. That's that's gonna be a a really fun time. It's already fun.
Arvid:Right? It's already fun to have these things do things for you, but to have that done in the background and have problems be fixed before you even notice them, I that that is a it's gonna be a weird future for people who are traditionally trained software engineers. That's how I feel.
Taylor:Yeah. I know. It's different, but I think it's exciting. I I don't feel threatened by it at all, you know, like some people seem to. I think it's really an exciting unlock in productivity for developers and just gives you so much more power honestly as a single founder as well.
Taylor:Like, if you're bootstrapping a company and it's just you working on something, can just be so much more productive, especially on maybe the boilerplate work that is pretty boring that often is where a product kind of gets stuck. Know, like if you've done a lot of side projects, you know, it's basically a joke at this point, how many unfinished side projects people, everyone has, right? And you they usually fizzle out because you lose excitement during the non boring parts of the development cycle, and being able to kinda push through that with AI is is pretty exciting.
Arvid:Yeah. I I heard a lot of people, and I was doing vibe coding on the side on the old projects just by throwing them into cursor or Windsurf or whatever and so like, finish this. That's really how that works. And I love this. I love that code gets reactivated, resurrected even, and then some agent takes maybe takes care of it.
Arvid:Maybe it goes nowhere, but maybe it goes to a point where people can actually deploy it, and then something is built that wouldn't have been built before. That's really cool. Do do you think you're ever gonna hire, like, a full time AI employee?
Taylor:I could see that happening. I wouldn't be I wouldn't be surprised, honestly. I'd almost be more surprised if we didn't eventually.
Arvid:Yeah. Right? Like, it's gonna be a cyborg team at some point. Yeah. Pretty wild.
Arvid:Man, that's bizarre. Well, I'm glad you have a very sensitive approach to AI. One thing that when I think about, like, AI existing AI systems, like platforms that help you build code, I wanna mention one thing. I asked my my Twitter feed a question recently about what the number one thing is that stops them from using Laravel for the next indie business. Because my my audience is a lot of entrepreneurs and a lot of software people at the intersection, and I would just wanted to see what's the vibe around Laravel and why do people not use it?
Arvid:Because the people who do use it, there's a lot of them. They are very happy with it and they understand that the ecosystem is really cool. The integrations are easy. If I want payment, I just do one composer require and I'm done. That kind of stuff is just extremely powerful.
Arvid:So I wanted to see the other side. And besides all the people saying, well, PHP sucks, what yep. You will find those people still, like, that are kind of left in the the PHP four point o world of 2005 or whatever, And people who still use JavaScript or they use JavaScript for everything because that's also, like, from the server to the front end to mobile. You can use it on every platform. Fine with them.
Arvid:The interesting ones were and I'm gonna quote this verbatim because I wanna see what you think about this. I don't use Laravel because the the fact that neither Cursor nor Replit ever suggests Laravel as an option. I found that interesting. Like, what are you gonna do about that? Like, how can you actually get your your framework into these tools?
Taylor:I mean, I think that's something we are working on that internally, even with the AI providers, you know, trying to improve this. I think it's like a valid point, right? And it goes for all new frameworks that will ever be created, right? It's created this weird situation where if I were to write Laravel today, it would like, the AI would have no idea how to write it. It feels like almost so much harder to launch an open source web framework today than it was ten years ago because of this very problem.
Taylor:And I'm very interested to like solve it because I think Laravel is uniquely suited to I think it's the most like expansive and cohesive full stack ecosystem you can write a web application in, in the sense that there's a solution for pretty much every problem you're gonna have more or less. And so for AI, just to be able to write all of that and you can deploy it, you don't have to go sign up for eight different SaaS services to sort of build your full stack app is really cool. So yeah, we're aware of this and like very interested in trying to solve this as best we can, because I think it's totally valid, like it's like the new, it's the new SEO, basically. You know?
Arvid:It really is. A a I o or or something. Right? Like, that's that's where we are. Yeah.
Arvid:I think Stripe has been doing something really interesting with her with their docs. Like, they it's almost like now native markdown, and a copy to markdown button is everywhere. And even, like, open this in ChatGPT is a is a button, like a button that's bigger than other buttons in their documentation. I think they've understood that they need to feed these things into the systems either by usage or by asking ChatGPT to crawl it so that it gets ingested into the next batch of training data for the model. That's the AI optimization stuff that we are still all just trying to figure out.
Arvid:Like, llms. Txt to describe a full website somewhere and then all documentation with code examples to just ingest. It's it's a bizarre world we're in because I didn't have this on my menu for 2025 that I have to export all my docs as a markdown file for some computer to eat. Right?
Taylor:No. I know. It's crazy. I never would have expected it, honestly.
Arvid:It's it's been a wild couple of years, but I can already feel, like, personally, how my software development experience is completely alien to what it would have been, like, two or three years ago. And that's not even with agentic stuff. Just even being able to take a piece of code, throwing it into some window and telling it, hey, fix this, and it does fix it. That is magical.
Taylor:I I laugh sometimes because over the years, you know, six or seven years ago, I would routinely ask like on Twitter, what should we build next for Laravel, you know, just to get feedback. And there was always these like joking responses of like, can it build my app for me? Can it like code for me? And now that is like a 100% reality, which is so mind blowing. And at the time, it just felt like a joke, you know, and that was not that long ago.
Arvid:Yep. I'm I'm not surprised that people are afraid for their jobs or just even for stability. I hear it from investors too. People who are investing in SaaS businesses, they don't know, like, where is this gonna go? Are we investing in something that somebody could build in twenty minutes, like, a couple years from now?
Arvid:Or is there a mode? What is the mode? Right? Like, the that is that is problematic. I I think what you're doing in building the the whole system, like, from the framework to the tools used to deploy it, to the place where it's deployed and the service, like the customer success teams that help you deploy this, I think that's the moat is to have control over the whole, like, channel from start to finish.
Arvid:But I can see how other people don't have that luxury. Right? It's like that's a real problem to them.
Taylor:Yeah. And for many people, the ideas were the hard part, in my opinion. Like, you can you can build apps that have no product market fit faster than you could before. You know what I Yeah.
Arvid:That's right. That's the one thing that that never changes. Right? Like, there's a there's a lot of bad ideas that are now just executed faster. That's yeah.
Arvid:That's pretty
Taylor:much it. Yeah. Which is fine. Better to, like, fail fast, you know, than to waste a lot of time and fail after years of develop hand coding Onto the next idea.
Arvid:Yeah. I I wonder where the balance is. I I do wonder where the balance is here because sometimes failing fast is, like, what, two days of no customers and the idea is bad. Right? Like, is should I take a month to to see if I can actually do some marketing?
Arvid:That that it might be too fast. And a lot of people have, like, AI FOMO and they just feel like they need to add AI features and whatnot, and they fail to see, like, what the actual value is.
Taylor:Yeah. Maybe it's screwing up our entire, like, business founder attention span of, I didn't hit a 100 k, you know, a week in the twenty four hours. This is done.
Arvid:Know? Honestly, if if you look at Twitter, that's exactly what's happening. Like, because the algorithms then also amplify these totally outlandish messages of people who had a fluke success or just random chance. Yeah, that's going be a problem. But I guess that's a problem for another day.
Arvid:I think we talk about stuff that's happening right now here, and I think Laravel is happening right now. To to anybody who thinks PHP is outdated in in 2025, what is your perspective as somebody who's built an empire on top of that language? You know, I think
Taylor:PHP has such an interesting history from being created in a very similar way to how Laravel was of just like a guy's personal collection of tools to build his own ideas faster to powering facebook.com and many other, you know, very notable startups. I think PHP is it's one of the few languages that was built specifically for the web and only for web programming in its original inception. And it's honestly just still such a fast, productive, no nonsense kind of language that anyone can learn. It's very quick to learn, even in just like a week, know, especially with AI assistance. It's not a very complicated language, but yet it's fast and there's tons of documentation, there's tons of educational material about PHP, about Laravel.
Taylor:And I still think it's really like the most productive way to build a web app, especially paired with tools like Livewire or Inertia where you can pair like a modern React or Vue front end with your Laravel backend in a way that just doesn't make you want to pull your hair out. I think it's a great way to build applications. And so that's why I still like, if I was starting a new business today, I'm not even thinking about anything else. Obviously I built Laravel, but I just think it's just the fastest, most productive way to build things. And I hear so many times of like, I'll hear developers venture into other ecosystems, whether it's like full stack JavaScript, or something else, and then come back and be like, Oh my gosh, like, I'm so much more productive on this stack.
Taylor:My team is more productive. Because we're not, you know, it's nice just to have a structure and a set of opinions on how to build things and not have to, like, reinvent the wheel for every possible thing you need to build, which I think is one area that Laravel excels.
Arvid:Yeah. And and I think our our community, I'm I'm just saying our because I feel like I'm part of the Laravel community, is not as fragmented and doesn't have as much infighting. There's not this kind of framework versus framework stuff. You literally have Vue. Js and React and whatever you want to put there.
Arvid:Swelter, sure, why not? You can find a way with inertia to to get all of these things working together. I think Laravel is is a is an integrated framework and and not like an excluding kind of framework.
Taylor:I I think it's very integrated and very cohesive. And I think what separates Laravel from maybe other, let's say more traditional full stack frameworks, is we're also I think an open minded community in terms of as developer tastes and practices change, we adopt them in a way that is not like reactionary, but like considered. And I think Inertia is like a great example of that, of like, look, it's just unrealistic to expect people to write their front ends using PHP echo tags forever, react in a large way as kind of one, let's say front end mindshare among developers. And we would be silly to ignore that. Similar, we would be silly to ignore AI and say, that's for the kids, that's stupid.
Taylor:We only handwrite artisan code over here, you know. So I've always been one to not reject where things are heading or not like blind myself to it. And we've tried to just kind of embrace, know, as the web development landscape evolves, we evolve with it as best we can and continue to try to build the most productive way to build apps.
Arvid:I'm not surprised. I think that's that's what PHP was too. Right? PHP was this collection of things, and then it got better and better with every version and integrated object oriented stuff and it moved away from these weird arbitrarily named things to a better internal structure. The time I built PHP was April somewhere in 2002 or 02/2003.
Arvid:I was doing TYPO3 backend extensions, and it was a horrible thing to do, but it worked. And that's the last time I used PHP prior to getting into Laravel, like a couple of years ago. So I had all this stuff still in my mind about how unusable it was, but I find myself not even using any of the functions that I used back in the day. That's how much the language has changed.
Taylor:Yeah. It's pretty crazy. And it's easy to, like I mean, you can download, a reactor view or LiveWire starter kit and just kinda look at the code, it it is pretty different. You know, it's very different than how PHP used to be written.
Arvid:Yeah. For sure. And and I think that's the great part. Like, you you it's a it's a thing that adapts to the reality in which it operates. That is it's the opposite of having this kind of very agenda driven framework development.
Arvid:I I don't wanna hate on anybody. Like, do whatever you want. I just love that this is the choice that I can now make and build my own monetized applications faster and more reliable than I ever could before. So I really appreciate you spending, like, your whole life building this, so thank you so much. It's really cool.
Arvid:And I I also appreciate all the the products you built along the way because not only are they helpful to me as a developer, they are inspirational to me as a founder. That's that is something and and to a lot of people out there. Right? Like, the the way you approach business and the way the decisions that you make along the way, what you've just voiced over the last half hour or so, that is why this is successful. So thanks so much for for sharing all these very amazing insights.
Arvid:I I can't wait to see what the future holds for Laravel. Where can people follow you and the businesses that you run and the journey that you're on if they wanna see more?
Taylor:Yeah. So they can follow me personally on xx.com/taylorotwell. You can follow laravel at x.com/laravelphp or just at laravel.com. And cloud.laravel.com is where you can check out, you know, our latest cloud platform and nightwatch.laravel.com will be coming soon. So you can get on the waiting list for that.
Arvid:That is amazing. Well, thanks so much, Taylor, for being on. That that was great insight into a framework that I love. So thanks for everything, man.
Taylor:Yeah. Thanks for having me.
Arvid:And that's it for today. Thank you so much for listening to The Bootstrap Founder. You can find me on Twitter at avid k a h l. And if you wanna support me in this show, please share podscan.fm, my SaaS business, with your professional peers and those who you think will benefit from tracking mentions of their brands, their businesses, and names on podcasts out there. PodScan is a near real time podcast database with a stellar API.
Arvid:We have 32,000,000 podcast episodes in the back now. The database is humongous. Please share the word with those who need to stay on top of the podcast ecosystem.
Creators and Guests
