The Laravel Podcast

The Laravel Podcast is brought to you by Tighten,

your friendly neighborhood Laravel experts.

Check us out at tighten.com

and also by other sponsors you will hear about later in the episode.

All right, welcome back to Laravel Podcast
Season 7.

I'm your host, Matt Stouffer, CEO of
Tighten.

And in this season, I'm joined every
episode by a member of the Laravel team.

And it's kind of weird to call Taylor
Otwell a member of the Laravel team,

but whatever. Today I'm joined by Taylor
Otwell.

What is your official title? Are you CEO?

I forget every time. Okay, CEO of Laravel,

Yeah, CEO of Laravel yeah.

OG, originator, creator, ~ papa bear,

whatever you want it to be. ~ Taylor,

would you say hi to the people?

Hey, what's up everybody? Good to be back.

Dude, it's been so long. you you know
this.

The listeners don't know this,

but I am ~ wrapping up this season and
planning next season.

And as I'm doing that, one of the things
I've done is like kind of gone back

to the previous seasons. And I have so
much nostalgia for when,

you know, both when it was just you and
me,

but also when it was ~ you and me and
Jeffrey and when we had Adam stuff,

Yeah. Yeah.

I'm just like, ~ we've had so many good
times on the podcast.

So anyway, thanks for hanging out.

Yeah.

Yeah. It's good to be back. It's good to
be back.

It's been a minute.

Okay, so there's so many things I want to
talk to you about,

but Laracon is next week, and I think I
remember Kayla telling me that this
episode

that we're recording on July 20th,

I think this is coming out on like the
21st or the 22nd.

So this will come out before Laracon.

Okay. Okay.

So I know that you're in Laracon prep
mode.

So what's going on? What's interesting?

What's new? What's different? If people
are you know aren't going or are going,

what should they look forward to?

Kind of what's what's the situation with
Laracon right

Yeah, I mean it's it's we're in the final
stretch here,

the last week. ~ I'm really excited about
this year's Laracon.

I think the venue's really cool.

I really like Boston as a city,

honestly. We went there as a family,

~ I guess. Yeah, last fall.

Yeah, last fall. Really enjoyed it.

Cool place.

Did you guys

know it was a potential venue when you
went?

Don't think we did actually, but it was
soon after that.

Okay. huh.

It was soon after that that we got that
figured out.

~ no, but I I'm excited about it.

I mean we have a lot of great talks.

I can speak to my talk a little bit.

You know, I'm gonna be doing kind of the
usual ~ demo of some like open source

Yeah.

stuff, just like in the code, like I
usually do.

If you've ever been to a Laracon I kind of
have like a routes file,

I show some new features, we kinda walk
through it together.

So there's a chunk of that.

~ and then I'll dive into ~ some of the
Cloud stuff we've been doing.

And I should say on the open source side,

you know, there's a mix of kind of like
some framework stuff,

~ some inertia stuff, some AI SDK and
boost stuff,

actually some pretty cool stuff coming
there.

~ and then I'll dive into some Cloud stuff
and kind of a recap

of stuff we've recently launched,

maybe in the last like month or so,

as well as ~ some things that are
launching sort of like now at Laracon.

And then ~ a preview of some things that
are coming in the next couple of months,

let's say. So it's a it's a it's yeah,

Yeah, you've been teasing some things on
on Twitter that I'm

yeah.

I'm I'm very excited to see what you're
doing there.

Yeah, I mean let me look at I've got the
list of my Laracon topics pulled up.

One, I think, four, five. There's at least
probably like eighteen

or nineteen things. I'm probably gonna
have to prune it a little bit.

My god.

~ so we'll see.

Nobody's gonna complain if you

go long. I I'm gonna tell you that right
now.

I don't know. I I assume are you like at
the end of the first day like usual?

Okay. Yeah, if we have to go late because
we're listening to Taylor Otwell talk

Yeah. End of the first day.

about amazing new stuff. Maybe the venue
will complain or something,

Yeah, yeah, yeah.

or whoever's running ops, but we won't
complain,

Yeah, I feel like

so

last year's keynote was a little long,

so I'm not gonna try to go that long,

but

Well, I I didn't notice it, but I know
that when you you have a space booked
they're

very sticklery about kind of when you wrap
things up.

But from our perspective, give us the
Taylor man.

Yeah, yeah.

Like okay, so there is new and and I have
actually tried to have multiple people

Yeah.

from the team on the podcast, and they've
literally told me there's nothing

I can share that will be even remotely
interesting because everything I've worked

on is top secret, but you can have me on
after Laracon if you want.

So I'm like

Yeah.

Well well damn So I'm very excited.

Yeah. Awesome.

Okay, so I've never been to Boston,

but I know it's at the Sowa Power Station.

So it's this like big kind of brick
building that's historic

and beautiful and everything. But I assume
you've kn big because you guys

did your vacation that did you fly back
and actually see the SoWa Power Station

in person once you guys picked it,

or have you never so you've never been to
the venue?

my god. It's a different world.

actually that ~ I have not seen.

I saw last year's in Denver. ~ well,

Uh-huh.

did I? I don't even know if I saw that one
beforehand,

but I definitely have not seen this one in
Boston.

~ so I will be you know, surprised myself.

I actually probably also know the least
about sort of just like what

That's awesome.

the stage decorations look like and things
like that than any previous year.

So I've kind of left that to Laracon team.

How do you feel about that? Is that a
peace

is that a peaceful thing for you?

Yeah, I'm fine with it. It feels good.

I mean, it'll be fun to be surprised a
little bit.

Plus, ~ the Laracon team, th they've done
a Laracon or two before,

so they know, like, you know, what to do.

So

~ so ~ this is not live as of the
recording,

but it will be live as of publishing.

~ and you already know about this,

but I built a a thing called Voxacon
Anyone who hasn't seen it,

just go check out Voxacon V-O-X-A-C-O-N.

Laravel.cloud. ~ and one of the things
that it is is a as accurate

as I possibly could get in a Minecraft
style thing representation of the the
venue.

So I have been poring over

every blueprint, every picture,

every video I could find of the SOA power
station.

So it's crazy because one of the things
that I'm very curious to see is when

I get there, like what did I miss?

You know, like what was not accessible.

There's one thing I know for sure.

There's no pictures of the kitchen in the
entire internet.

So I have no idea what the kitchen looks
like.

~ but outside of that, I feel like I have
a more intimate understanding

of the layout of the Sowa power station
than any other person in the history

of the world who has not been there in
person because I've been obsessing over
it.

That's funny. That's

Yeah.

funny.

Okay. So you mentioned one of the things
you're going to be introducing

is Laravel and AI. And I wanted to hear a
little bit about ~ you know,

like we have been talking a lot publicly
about how Laravel is an ideal

~ environment for somebody who wants to to
work in AI.

And I think that's true for those of us
who are already,

you know, writing Laravel. You've got the
AI SDK and stuff like that.

But I think part of the kind of the case
that we're making here is also that it's

valuable for people who don't currently
use Laravel.

So I live in Atlanta. There's a lot of
people in Atlanta Tech Village

Mm.

and and you know the surrounding tech
world who are doing the startup world

and they want to find a productive
ecosystem for developing applications,

using AI, ready to ship securely,

stably, as quickly as possible.

Can you give the pitch for like what have
you been working on with Laravel

to make it the ideal environment for those
types of folks?

Yeah, I mean, I think, first of all,

I agree that like Laravel works really
well paired with coding agents.

And I pretty much see that every day,

you know, in my own work with Laravel.

I think the foundation for a lot of the
work we do around making Laravel great

for coding agents is with Laravel Boost,

as far as like new package development
goes,

we've been working on some new boost
features,

which I'll show at Laracon that make ~

cool.

you know, building Laravel apps even
better.

~ but you know, Boost basically feeds a
bunch of context about Laravel into

the coding agent, gives it access to the
docs,

things like that. ~ I think we also just
benefit a lot from Laravel's opinionated

structure of as we've said, you know,

many times before, where the agent sort of
knows what to do.

It knows how to access the database.

It doesn't have to like, you need to go
sign up for Superbase,

you and and link that into your project.

Yeah

Yeah.

not that super base isn't great,

but you like the database is just
available,

you know, right there. and just all of the
conventions,

you know, around Laravel I think make it
really just a productive stack

for like hitting the ground running with
minimal sort of like setup

and decisions you have to make.

Yeah.

You can kind of just like start almost
vibe coding right out of the box,

you know, with like a complete package.

Yeah. Yeah.

~ you know, so I think that's like the
main thing that makes it.

really great. We are working on a couple
of other things,

~ which I'll show at Laracon that I I
think will make it even better.

Some boost related, some not boost
related.

~ but yeah, I mean I I'm actually
continually surprised.

And during my prep for Laracon,

it's been interesting because like
historically one of the most time
consuming parts

of prepping for Laracon was like prepping
the demos themselves,

especially if it had like a graphical
component or was like a dashboard where

I needed I needed to show off some
feature.

And now with like I was using Boost and a
fresh Laravel app.

Yeah.

And I was like, hey, I want to demo this
feature at Laracon.

Here's kind of like the docs for it.

I need to build a dashboard that has these
buttons and lets me interact with

it in this way. And it's just like done,

you know, in five minutes. And it's
basically one shot it.

And it actually used the inertia react
like boost skill to like know what to do.

Yeah, that's cool.

And that just that would have taken me
like all day pretty much,

you know, to to build that kind of thing.

And I've

I would have made mistakes and like had to
fix stuff and hunt bugs down

and like it just so much time saved.

Yeah.

Wow. I mean that's you you've often said
in the past that one of

the things that makes you able to build a
framework that is easy

for newcomers is you constantly put
yourself back in kind of this newcomer
state.

And I mean, I don't know about you,

Yeah.

but for me, every time I you every time I
hit Laravel New when it's been

a few months, I feel like I'm able to
enter that newcomer state.

I'm like, What what what

Where do I do that? Where is that?

And the easier that process is,

and to your point, it's so much easier now
than it has been in the past.

The more I'm just like, okay, like I can I
can walk a newcomer in.

If I if my brain, who has been maintaining
legacy apps or doing marketing

or whatever for the last year all of a
sudden has to start from scratch,

I'm getting some of that newcomer energy.

And I imagine these kind of like these
prep experiences are like that for you.

Yeah, definitely. And and it's it's crazy
how easy it is to get started,

~

you know, these days, just in programming
in general,

but especially with Laravel, to where,

you know, a couple of years ago someone
told me they wanted to like learn Laravel

and start building apps, I would have
like,

my gosh, like this is gonna be like yeah,

Where do you start?

this is gonna take years, you know,

to like get up to speed. but now it's just
like chat with a coding agent about it.

Yeah.

If you get confused, just ask it a
question and keep moving along.

Yeah. Yeah.

Well, I mean, you and you and I both know
Kayla Helmick and she was featured

on the the Laravel blog. For anybody who's
interested,

we'll make sure that she's linked in the
show notes.

And she is has no history as a programmer,

~ but understands programming between
being friends with you and and being

my VA and stuff like that. And ~ you know,

Mm-hmm.

she's built multiple functioning Laravel
applications from scratch

Just using an agent. And she knows her
bounds.

I mean, not everybody does. So one of the
things she knows,

she's like, I'm not going to do anything
that takes credit cards.

I'm not, I'm not even doing anything that,

you know, takes usernames and passwords,

but still super functional tools for her,

for her, ~ for her clients. And that's
been really transformative to me

Mm-hmm.

to watch her experience. Cause this is not
theory.

We're not just throwing out phrases,

batteries included this or whatever.

It's like, no, we have a real person we
know who doesn't understand

a line of programming who has built
functional applications that clients have
paid

her for. And I'm like, that's different.

Yeah.

Yeah, it really is. One thing I also
notice,

That's really special.

you know, with Laravel at AI is I'm
starting to see or or maybe this is

one reason I would say that Laravel's
great for even new people that

are interested in building software.

I've been getting more and more sort of
like a lot of it is just through real
world

friends, so called real world outside of
like Twitter,

you know what I mean? Like people I know
in Arkansas.

Yes, yes.

Yeah.

that are either referring to me people
they know because they know

I do web development and I'm seeing more
and more people that are using sort

of like vibe coding platforms that are
getting like call it fifty or sixty
percent,

Mm-hmm.

maybe even higher, of like on the way to
their idea.

And they eventually hit sort of like a
wall or a limitation in

the platform or something, or they run out
of tokens b or whatever,

and they're like, I'm stuck.

And I think Laravel is really a great
alternative to that because

Yeah.

you don't really have the same ceilings,

you know, like we when when you're using
Laravel and Boost,

you can build even like multi tenant,

multiple users, ~ using the app.

You can just get a lot farther,

a lot cheaper usually, honestly,

~ by just by linking it to your like
existing chat GPT account and using

a coding agent. And you just don't have a
lot of the same barriers

~ that just people seem to inevitably hit,

you know.

on sort of like the prepackaged vibe
coding platforms.

Yeah. Well, and there's something fun
about that too,

though, is we have taken now several
clients from a lovable or replit

or whatever else vision to a production
scalable secure Laravel

application. And we're able to do the
final build for them faster than

we would if they had not used those vibe
coding platforms.

So there's actually a way where I'm like,

Mm-hmm.

regardless of whether you started with a
lovable or not,

launching on Laravel still makes sense.

Right. So you either start from Laravel on
s on day one,

Yeah.

that's totally fine. But like for example,

we had a client who's just like,

he's like, I don't have the budget it
would take for you guys to build this
whole

thing in in Laravel, part of which is
because I have not fully defined the
thing.

And I know that there's gonna be a ton of
back and forth,

and I'm not gonna be a great client.

But if I'm going back and forth with
lovable for six months,

then when I get the lovable to the point
where it looks good.

Then I hand you the lovable and you
actually make it functional on Laravel
base.

And we did, and he launched and he's got
paying subscribers

and everything went great. And I'm just
like,

and this has happened multiple times.

Yeah, it's crazy.

So I'm just like, Lar Laravel should be
the end result,

regardless of whether you started Laravel
or started something else,

because there's no way that he could have
done,

he could have hosted on Lovable or Replet
or whatever else the level

of complexity that he needed. But we could
build it in Laravel faster because

Yeah.

Lovable built the whole front end for him.

So it's it's really fun to be able to say,

Yeah.

regardless of where you start,

this is where you should land.

Yeah. Yep. And it's it's crazy how like
using AI,

a big mindset from shift for me has been
that like code is sort of like cheap

in a way. Where like even if you build
something on Loveable or

Mm-hmm. Yeah.

you start somewhere else and you're like,

I kinda wish I would have done this in
Laravel,

you know, previously that would have been
like,

ugh, like this is gonna be a pain,

Rewrite Hell. Yeah, yeah.

you know. Like Yeah. But with with with
AI,

that kind of stuff, like the code is not
so

intimidating or like, you know,

those kinds of migrations are maybe a
little bit more palatable because like

it is so much faster to do with AI.

And like, you know, I've had to like train
my mind to think that

way where it's like, ~ I want to try some
kind of quirky big idea.

Previously that was a lot of commitment,

you know, like, do I really want to devote
eight hours to exploring this?

Yeah. You gotta really know that it's
worth the time here.

Yeah.

But now if I can kick something off in
five minutes and just get a sense

of like if it's a good idea or not,

that's so much time saved.

Yeah. ~ have you found yourself in
positions where you generate ~ stuff that
you're

like feeling exhausted from having to
maintain it,

or are you like, no, I just feel
comfortable throwing it away,

it's no big deal.

In regards to like Laravel framework,

or what do you mean?

Well, just any so one of I've found the
same thing as you where it's

I'm able to generate more experiments than
I have in the past.

But sometimes I'm like, ugh, I spent all
this time,

Mm.

all this time, because really like I spent
this time prompting this thing.

I told people about it, I launched it,

Yeah.

and I'm like, you know what? This would
not have stood the test of is this worth

my time in the past? And it wasn't worth
my time.

And I've just had to give myself
permission to say,

That's fine. It was an experiment.

I launched it and now I'm deleting it.

You know, like and I just tell people
internally,

Yeah.

I'm like, Hey, you that internal tool I
built?

I'm gonna get rid of it 'cause it wasn't
worth it.

You know, have you ever had things where
you're like,

Yeah.

I've gone further down a road than I would
have before 'cause it is so cheap

to do so.

yeah, totally. And I think it's part of
the training my brain is that it's okay

to delete it. You know, like i I would
have held on to it a lot harder.

Yeah.

And I think my I want to sometimes because
of,

you know, it used to be that it required
so much time investment in code,

Yes.

but now like with the code kind of being
disposable in a sense,

like I'm more comfortable or I'm trying to
become more comfortable saying,

this didn't work out. We can just trash it
entirely,

you know. ~

Yes. Which is a weird

thing to learn because it would have been
so wasteful in the past,

right? And now you're like, look,

Yeah.

I it was a it was a thought experiment.

I and it was possi valuable for me in this
way.

I learned these things. And that doesn't
mean it has to be in production.

That's okay. Yeah. ~ okay,

Yeah, totally.

so I before we step off of AI,

there is one thing. We use the phrase
batteries included all the time,

and and people a lot of team

Mm.

people marketing Laravel out to the
outside world talks about Laravel's value

as a batteries included framework,

even before AI. Can you talk a little bit
to somebody who's been hearing that phrase

and isn't actually familiar with what are
you talking about?

What does that actually mean and how does
that set Laravel apart from other

frameworks?

Yeah, it's a good question. ~ when it
comes to sort of like popular

web frameworks in today's age,

there's kind of ~ I I mean the most
popular framework,

you know, in the JavaScript sp space,

which is Next.js, doesn't have a lot of
opinions on how you do various things,

which is can be good or bad. So like it
doesn't necessarily have strong opinions

on how or where or with who you run a
database.

~ it doesn't necessarily have super strong
opinions out of the box on

how you do authentication and things like
that,

or how you do OAuth or social authors or
things like that.

So what it means is that like if you were
to start building

an application like that, or using other
popular like frameworks in that space,

there's a lot of sort of like pre
decisions you need to make that maybe
you're

not even really equipped to make ~ around
like where what kind of database

Yeah.

Where's your auth gonna be? How's it gonna
work?

How are you gonna do like scheduled tasks?

How are you gonna send emails?

~ you know, all of that stuff you're gonna
have to make decisions on,

or the agents is gonna maybe make
decisions for you,

which is probably likely the case.

and it can lead to a little bit of just
like vendor sprawl where

you're signing up for a lot of different
services,

you know, for across database,

auth, email,

maybe even Q jobs, you know, it's just
spreading out quite a bit.

~ it's also gonna be like there there's
not necessarily a structure or

an opinion on how the code is actually
architected to do those things in many
cases.

So it can lead to a little bit more of
like,

Yeah.

you know, we know AI is great,

but it doesn't always really like
architect the most like firm foundations

Yes.

for a long lived project that's you're
gonna want to maintain for years

Yeah. Yeah.

and you need to be stable.

it it can build a demo really quickly,

but as far as like longevity and
maintaining it over long time horizons
still

requires like good architecture,

some a lot of times human input and kind
of human review to make sure it's

not going off the rails. ~ so with
Laravel,

a lot of those opinions are baked in.

A lot of the even the structure is baked
in of where certain files should

go and kind of what should be in them and
how they should all work together.

And those opinions have sort of been
formulated.

over many years of building web
applications ourselves,

but also just kind of learning from what
came before Laravel,

Yeah.

whether that was Ruby on Rails or other
popular frameworks at the time,

and sort of incorporating the best
practices that generations of

web developers have sort of distilled.

~ so that leads to sort of just like out
of the box,

kind of more maintainable code that's
gonna just like survive longer

and hold up better.

That's not to say you're never gonna need
like human eyes on

AI generated Laravel code, 'cause you
probably will.

~ but I think it's just gonna get you
further faster and without necessarily
needing

to result to signing up for a bunch of
different vendors,

you know.

Yeah. Yeah. I mean, with our clients these
days,

you know, the biggest complaint often was,

I have to sign up for Forge and
DigitalOcean.

So like, well, those are pretty easy
infrastructure changes.

Mm.

So you know, just sign up for Forge VPS or
just sign up for ~ you know,

cloud or whatever. ~ but it is definitely
not a point where you don't have

Yeah.

a managed queue provider and a separate
database provider and

all these other things. ~ but I also
remember onboarding some JavaScript
developers

into Laravel I don't think it was ten
years ago,

but quite quite a while ago, and they were

They were a little bit overwhelmed because
they said there's so many files

and directories. And in some ways,

you know, you've worked on that to say
there's less off the bat.

Mm-hmm.

But part of what I had to tell them was
part of the reason there's more here

is because it does more things,

right? Like what you're seeing is stuff
that's actually functional out of the box.

Yeah.

And yeah, there might be more things here,

but it's because you don't have to build
those things.

And bike shedding is a real problem,

man. Just like, we're we have to start
something.

Now let's spend the next six months
deciding.

And and on every app you've ever built,

it's different.

So good luck remembering how to fix it and
having you know,

and now have good luck having the AI know
how it works because first thing

Yeah.

it has to do is research which of
seventeen different situations you

set up with versus with Laravel It's like,

I know, I I know you're using eloquent,

so it's gonna be eloquent, so that's it.

Yeah.

Yeah.

Yeah, for many developers the bike
shedding part's the part they enjoy the
most.

Yeah, so let's talk about that

because I I've you you just mentioned
something about code being kind

of ~ throwaway, but you're also the the
eloquent artisan guy,

right? Like you, you know, an origin story
of Laravel is,

I know. Well

you know, what does it make sense if I pay
extra attention to my,

you know, my lines and everything like
that.

So how do you how do you feel about saying
that?

That like is it some are you disappointed
in any way?

Cause I know we've had a lot of
conversations online about like some
people

it was always about writing the code and
some people it's about building

the product, but you've always been both.

So what does this new world feel like for
you?

I think it is it's a great question.

I think it is a very like interesting time
where like,

you know, historically Laravel and myself
have taken a lot of pride

in sort of crafting these sort of really
elegant APIs and interfaces

for doing different things, whether that's
just being able to say like

Yeah.

Q colon colon later, you know,

to to delay a job or something like very
concise,

very memorable syntax that sort of is a
pleasure to read.

I still do that for what it's worth on all
of the open source stuff.

So we just worked on a really big feature
for the AI SDK,

which I'll show at Laracon. And ~ you
know,

a lot of the code is like AI assisted and
generated,

but I still reviewed every file.

I still sort of like sweat the details on
how the user is gonna interact with this

API, even if

No user ever does. Maybe only AI generates
this code,

Yeah.

but like I still spend the time thinking
about like,

okay, let's assume that users are going to
be interacting with this.

Like, what is the most like intuitive,

joyful to write API? Maybe that's like a
lost cause,

but I still kind of derive like personal
satisfaction from that.

I think a lot of developers actually
probably still do.

Like, even though the AI is assisting you,

you still find joy in sort of finding just
the right

abstraction or just the right name for
something,

Yes.

that is still like a really fun for me.

And again, I still just derive personal
satisfaction from knowing that sort

of like you're building like a dresser,

right? And the backside of it is still
like finished perfectly,

even if no one ever sees it because you
know and you know that the tool

is like good and well built. And for me,

like I still don't want to use tools that
or I I wouldn't really

be comfortable using Laravel if I didn't
know

that the whole thing was thought through,

you know, all the way through.

Yeah. Yeah.

~ so I I still do that and maybe that will
become sort of an outdated mindset over

time. It's possible, right? As AI keeps
getting better.

But for now I still find it personally
satisfying.

And I I it's one that I hope that you
continue to take the luxury to do,

~ because A, I think it's just good and
beautiful to do that.

But B, ~ I'm still one of those sticklers
who says,

even if I'm using AI, I'm gonna read every
single line of code

and understand every single line of code.

And I've found a draft drastic difference.

Yeah.

As as I have been using AI to delve into
worlds outside of Laravel that

I wouldn't have done as much in the past,

I find a drastic difference in the ease
with which I can

Comprehend and review Laravel code and
code outside the Laravel ecosystem.

So, like, even if the AI is writing it,

Mm-hmm.

if I'm reading it, especially if I want to
have to skim it,

because whatever, like I can, it's so
expressive that it's just very clear what

the Laravel code should be doing.

And that's not the case everywhere else.

So e if there's a human in the loop,

Yeah.

even if the human didn't write the code,

I still think this human-friendly

Coding is going to like if if I were to
look at two two sets of code that were
spit

out by, you know, an AI, ~ I'm much more
likely to comprehend

and therefore continuing to choose to
read,

interview, and comprehend and take
ownership over the Laravel code than
something

Mm-hmm.

else where I'm like, I don't understand
this crap,

whatever. And once you hit that whatever
point,

you've lost your ability to, you're never
gonna go back and learn it more once that

hit you first hit your whatever,

right? Like it's gone. Yes, exactly.

So I think it's beautiful just for that.

that has has gone into the Laravel
framework,

you know, in the last what is it,

maybe seven or eight months since sort of
agentic coding really exploded.

I've still reviewed every line that has
gone into the framework by hand,

Love it. Yeah.

essentially.

Yeah, I love it.

So I want to thank some of our sponsors.

First, I want to thank our sponsor,

Laravel, the wonderful folks behind
Laravel Cloud and Laracon US.

Laravel Cloud is the fastest way to deploy
Laravel apps,

and it also solves all sorts of common
problems that keep

web artisans from shipping. And Laracon US
is their premier in-person event

for all web artisans to come be with your
people,

level up, and leave ready to ship better
and faster than ever.

Check these out at cloud.laravel.com,

laracon.us

Taylor and I, of course, will both be
speaking at Laracon We'll both

be there in person. And we are talking ~
both in our talks about Laravel Cloud.

So definitely check it out. And also
Laracon Australia,

where I also will be. I don't know about
Taylor.

Laracon Australia returns November 4th to
6th in Brisbane.

I love Brisbane with deeper learning,

practical insights, and ideas that will
stretch how you think about building with

Laravel. So get your tickets now at
Laracon.au

so you know one of the questions I wanted to ask

was kind of what's the future of
programming?

So let's let's go there because I think
it's a pretty natural transition.

Mm.

As you think about the future of
programming,

you have built tools for programmers your
the majority of your adult life.

You've built, you know, open source tools,

you've built paid tools, you run a
company,

~ building, you know, dev tooling and also
providing ~ open source tooling

for developers. ~

What do you think the future is of
programming?

And I mean that in two ways, and you can
add more ways,

but the two I mean for sure are the the
future of the the role that

we as programmers have, ~ and also the
future of the way that other people
outside

of us are doing programming, you know.

So, you know, for us is are we gonna be
writing code?

Are we all becoming product owners?

Are you know we're becoming solutions
architects instead of hand handled coders?

Mm.

Are other people outside of us now gonna
be doing a lot more code?

Like as you look to, you know,

a couple of years from now.

What is your sense of where it's heading,

or do you know any more than anybody else?

I sort of suspect that we're living in it
now,

~ to a large extent. And what I mean is
that like

I think that the main shift has actually
already happened.

~ in the sense that many of us are
actually not writing the majority of

the code by hand, maybe none of the code
by hand.

~

~ I think that was like the big thing.

I think like what remains maybe like to
come or is less clear is how much

of the review will we be doing versus like
trusting the AI.

And I think people are already at sort of
like various levels with this,

you know.

in different places. So like I actually
just went to Anthropic's office

for a founder day, ~ that they did and got
to hear from a few of the people working,

cool.

~ particularly on Claude Code,

~ but someone kind of on the research side
of Anthropics model stuff as well.

And I think they are very like high trust
of the model.

They're obvious obviously using like the
model

before we get it and at a faster speed
than we get it and maybe with higher
effort,

you know, and less token limits.

But they seem to have like very high trust
in the model.

And I think their their view that I
gathered out there was that,

you know, the future is not really
reviewing the code.

It's more of a agent manager.

You're sort of like guiding the agent or
steering it in terms

of the overall solution, but you're not
reviewing every line of syntax.

they were also very like goal or loop
focused of you know,

kicking off long running goals with the
agent that are maybe running for hours

at a time to do big projects. But yeah,

I mean I mean to come back to the
question,

I think the the main shift has happened in
the sense that we're not code writers.

~ I think that is like already done.

It's just a matter of like how much trust
are we gonna place in the agent

as far as the review and overall
architecture.

And I think it's hard to say. I mean,

if we if we keep going projecting out
years,

you could imagine with model improvements.

I mean, it's you could imagine a world
where the code is trustworthy.

You know, like it's not science fiction,

Mm-hmm. Yeah.

I don't think, to to think about it a few
years out to think that,

well, maybe it does become sort of like a
waste of time to review all of the code.

Yeah, yeah, yeah.

~ But you know, I don't I don't think
we're there yet,

but again, you could imagine a world.

And at that point,

I don't think it actually looks all that
fundamentally different from what

we do now. You know, I I think we're still
gonna have to have a basic grasp

of like the technologies available,

how they work, what the upsides and
downsides are maybe to different databases

or to different strategies for building
things or different overall architectures,

especially at scale. ~ but yeah,

I you know, I think ~ I don't know.

That that's the best I can like kind of
state it,

I guess, from where we are now.

Yeah. Well, I don't I don't know if I've
heard anybody say that before.

That that that we're there already.

Mm.

~ everybody's got visions, whether
positive or negative,

but ~ I I'm actually really heartened if
that's the case,

because so everybody at Tighten still
reads every single line of code,

and a large portion of the code written at
Tighten is handwritten.

~ I have told everybody you need to be an
expert in AI.

Like that's the baseline expectation.

Mm-hmm.

So they're all experts in AI.

but they use it to differing degrees.

They use, you know, different models,

although claude code is the most common.

~ but I have noticed that in the agency
world at least,

~ the people who are doing the best are
people like us,

where we are here for our expertise.

Like we're here because we understand the
architectures,

the systems, the decisions to make.

We're here because we've seen thousands of
code bases and because

we know what's going to scale and what's
going to be secure.

And those things are implemented through
us actually typing characters into an IDE.

But the characters in the IDE is not
really what Tighten brings uniquely
compared

to somebody else. That's just an
implementation detail of our expertise.

So the implementation shifting doesn't
change the fact that you hire Tighten

Mm-hmm.

for expertise, right? Whereas with people
where your job is to have

Mm-hmm.

an expert tell you, now go type characters
into an IDE,

those are the folks that are having the
hardest time.

And my hope is we're able to keep a
pipeline where people can be in that
stage.

And still be able to move up. Otherwise,

we're never gonna have any a junior
pipeline anymore,

right? So there's definitely things to
consider about there.

Mm-hmm.

But my hope is that we can find spaces for
people as they are going through

a junior pipeline to be learning maybe
earlier how to be expert thinkers,

rather than spending their first five
years just learning which characters

go into what's state in the IDE.

So

Yeah, I think that makes a lot of sense.

I think a lot of people if I mean I think
a lot of people are also like

Okay, yeah.

my AI coding setup is I think much more
vanilla than a lot of people

Yeah.

are thinking the future of AI coding might
look like.

I think I think I think Dax I think Dax
and Open Code had a great tweet about this

So tell us, what is your setup?

a few days ago. I mean my my ~ setup is
just like either Claud Code or open code

Uh-huh.

and I type into it and code comes out at
the other end.

Yeah. Yeah.

And there's

no like advanced orchestration layer
beyond that.

Yeah, yeah. Yeah.

Rather the in just whatever is built into
the vanilla setup is sort of what I use.

I think a lot of time is being spent sort
of like trying to I think this

is very like developer brained,

by the way. Like trying to like come up
with these more like fancy

Yeah.

and efficient ways to like have tons of
things going at once and like it's

Yeah.

a big software factory black box.

And in a lot of ways I feel like anthropic
actually does operate this way,

which is maybe like a counterpoint to how
I see it.

Yeah.

But I just see I just like think a lot of
that is sort

of another flavor of I'm gonna build my
own framework.

~ it's that classic like developer brain
of like,

Yeah.

I'm the special snowflake that wants this
special setup and I need

Yeah, yeah.

all of this abstraction around it to make
it work well.

Yeah.

Instead of just like using the tool like
to do something useful,

Yeah. Yes. Well

you know? Like and I I think that that is
what I mean by I think

the fundamental shift in software is like
with AI sort of already we're living

in the future of it. I've just that was
the fundamental shift is having like this

terminal or text box that I can type into
a plain English thought and code comes

Mm-hmm.

out the other end and multiple files are
changed and things are done for

me on my computer.

That that was like the futuristic shift
and ~ you know,

Yeah.

I think we're already there in that sense.

Yeah. And I I love that. My other podcast,

Pragmatic AI, I allow myself to get a lot
more into like what's happening this week.

And we just had John O'Nolan on and John
was talking.

And John has always been one of my friends
who is A,

building real things and B at the
forefront of AI.

I feel like there's very few people who
are both of those.

You're either at the forefront of AI and
you spend all your spare time building

Yeah.

yet another orchestrator, you know,

or you're building real things and you're
like,

Yeah, AI is helpful in some ways.

But you're not like

Just.

Again, trying to bleeding edge spend all
your time and energy on orchestrators.

~ so but it was he and he mentioned goals,

you know, he mentioned goals as a thing
that he's using,

and I hadn't even been switching to goals.

So I I'm I appreciate the fact that I
don't have to be in the latest,

greatest to feel really productive.

in part because the people who are and
that's one of the reasons why

I'm just like I'm perfectly fine if I'm
not doing the same thing as anthropic,

because the people who are at the absolute
forefront,

they will spend

Sure.

spin through things in two weeks and then
I would be like,

Yep, you you you were hype about it.

You talked about it nonstop for two weeks
and then it never ever came up again.

Let me see something that surfaces and
actually sticks around for three months

and then maybe we'll talk about it.

Yeah.

it remind it's a little bit reminiscent
also of like the you know,

like when Facebook was popping off in that
era of ~ startups

and there were microservices and things
like that where like the individual devs

or maybe the two or three person teams
were like splitting their app with

a hundred users into five microservices.

And it's just like th they're Exactly,

Cause AWS and Facebook do it.

because these sort of like hyper scaled
startups were doing it.

There is maybe a little bit of a similar
thing in AI where

Yeah.

You know, you would be better served by
just popping open the agent

and typing something into it and not
spending so much time sort

Yeah. Yeah.

of like orchestrating around it.

Yeah. Or maybe open up a second tab with
second agent or use solo.

Yeah. Yeah. Use solo. Shout out solo.

So you get you know, like it's just sort
of like Yeah,

love solo. The the most complicated thing
I do is if I want

to have distinct sub agents, I've found
using solo to do it is a little

bit easier than having the agent manage it
itself.

And that's it. Outside of that,

Yeah.

I'm just sort of like, Yeah, I've got an
agent doing this,

an agent doing that, I'm doing this.

That's enough.

Yeah, totally.

And to thank some more
sponsors,

move fast and fix things with Honey
Badger's intelligent error tracking

and application monitoring for Laravel.

It's everything you need to keep
production healthy and users happy.

Get your free developer account at
honeybadger.io.

And also, this is not an official
sponsorship,

but I do want to let you all know that at
Laracon US next week,

our friends at Geocodio and I are and we

We, Tighten and Geocodio, have partnered
together to give away 500 free

ice cream sandwiches from a local Boston
shop.

You just need to come by both of our
booths and get a little thing stamped,

I think it's gonna be. And if you take the
double stamped thing over,

you will get a free ice cream sandwich if
you're in the first 500 people.

So if you are coming to Laricon US,

please make sure to come by both of our
booths and get those stamps

and get some delightful free ice cream.

so one thing that has been really fun over
the last maybe two weeks

is your Twitter account. and Spicy Taylor
~ has lots

of opinions and it's been really cool
because I feel like ~ when everything
happened

with the investment and Accel and stuff
like that,

you you just had so much to manage,

right? You went from having a team at like
10 to a team of nearly a hundred,

and obviously you don't manage every
single person,

but as somebody who is really active on
the internet.

And then I took over as the CEO at
Tighten.

And then it was really hard to be active
the internet because I'm suddenly doing

all what I was doing before and now fully
running a company.

It's I've only been I've been pretty slow
coming back.

And so I've been watching you come back
and I'm just like,

like we've got Taylor again. If somebody's
not on Twitter,

d I could describe it, but I'm curious for
you.

Like, did you just feel like was there
just a moment where you felt free?

Did something change? Like what what is
leading us to the point where you're

all out you're thought leading,

you're sharing funny stuff, and I just
feel like we have old Taylor back.

Yeah. I don't I don't know. I I think like
I'm still trying to distill it myself.

I think like the last couple of years have
been a lot of like learning

Yeah.

and growth and lessons and kind of like
reflections.

I do think, you know, maybe I'm coming to
a place where like some things

are starting to click for me, ~ or I'm
getting comfortable with some things.

You know, but I mean just to go back,

yeah, like Laravel has grown a lot.

~ when we first raised a series A

In 2024, like you said, we were like eight
or nine people,

~ mostly devs. We brought on Tom and Andre
as kind

of our first like couple of hires.

and we started growing the company.

And, you know, those foundational people
had a lot of great experience that

I didn't have ~ and to some extent still
don't have.

I

You know, I am not I actually I don't know
that I really manage anyone at Laravel,

even today. I don't think I'm actually a
good like manager in the traditional
sense.

I collaborate a lot with people.

~ almost a little bit as like a peer,

even though I I'm I am like their boss,

you know, to a certain extent,

and we're not totally peers at the same
level in that way,

but I prefer to work in this like very
collaborative way.

~

So like I work a lot with like Joe Dixon
in in product,

~ where we're just kind of thinking about
like what are the projects we want

to work on next? Where where are we kind
of heading like long term?

~ and even that I'm not I don't I prefer
not to think like in terms

of multiple years. I think that's really
silly in tech.

~ if we would have tried to do that a
couple of years ago,

it would have all been useless.

~ but I try to at least think of c a few
quarters,

you know, of like where where do we wanna
go and kind of what are the like vague

goals we want to hit, you know,

in the next year or two. ~ I spent so I
spent a lot of time working on that

in open source and then just sort of like
collaborating across the business.

You know, it could be like in a marketing
thing for Laracon.

It could be on a website thing.

It could be even on an email we're sending
and like what the copy should be.

~ it could be looking at something for one
of our private cloud customers.

~ something's not working quite right or
in the framework itself maybe.

~ but you know like

Regarding sort of like the the the
reversion to sort of like old Taylor on
Twitter,

I think that like for me maybe one of the
biggest regrets

I have over the last couple years,

and not to say it was like a huge mistake
or it was a big regret,

was like not leaning into myself enough or
like not like really like

Yeah.

I don't know it's hard to say it without
sounding like proud or sort of like

conceded, but not sort of like just owning
that I have actually been successful

Yeah.

at the at like a software business.

Yeah. Yeah. Yeah.

~ in ways that other people of the company
maybe haven't.

~ I've actually, you know,

opened up an editor totally blank and
convinced hundreds of thousands

of people to use this tool. And the
business say millions of dollars,

right? ~ which is just like rare.

~ so I do think that ~

Yes.

I have a very like particular opinionated
way of doing things,

right? The spicy takes. Like you can do a
lot of things in one day.

Or like, why would you ever start with
like a longer time frame?

Yeah.

Or you could run the whole company out of
10 notion docs and 10 Slack channels.

Yes. Yeah.

I think that ~ I am very that is very true
to me as a person.

Like I am very radical that way.

So there is like so much I feel what I'm
saying is like actually true.

~

And just sort of like leaning into it a
little bit more of like,

what of I I think what I'm trying to say
is when we grew the company,

Yeah.

we brought out a lot of really talented
people in a lot of very various areas.

And I kind of like in some ways I feel
like retreated mentally

of like these people know these people are
professionals.

I am was more of like a hobbyist that got
lucky in many ways.

Yeah. I should defer. Yeah.

You know what I mean?

Yeah, absolutely.

And I'm trying to get

a get away from that a little bit to where
like these people it is true,

these people are professionals,

but I actually have done a lot of crazy
shit too.

Very successfully. And should lean into
those opinions a little bit more because

Yeah. Truly. Yeah.

I think like what I bring to the table is
like a nice counterbalance

to what other people bring to the table,

a little bit more like crazy, a little bit
more like swashbuckling,

you know, like just ship something now.

Like what if we just shipped it now?

Yeah.

Like what if we just hit the button right
now?

Like what would really happen?

Is it really that scary? You know,

like, ~ so trying to push on that a little
bit,

Yeah.

you know, push the pace a little bit,

push the energy a little bit. ~ you know,

which I think is like what I can bring to
the table.

~ and I think other people at the company
bring amazing things to the table,

but I think we all need to bring our
unique selves to the table.

And I think I was trying to be a little
bit more like,

blending in, you know, with what everyone
else brought to the table.

And that's really probably not what the
company needs from me,

actually. You know, we already have that.

And we need me to be me. So it was me
trying to like,

a lot of it's me preaching to myself,

you know, of like, I need to be like
pushing the pace.

This is true. Yeah.

You know, it is so easy like when you're
building a software company,

like, ~ especially remotely,

I think. It's hard to feel the like
intensity and the energy when you're all
remote.

And so like I do want to be someone at the
company that sort of is like always sort

of like, you know, like, okay,

what are we working on this week?

Like what are ex trying to like bring
excitement and energy to the to the space?

'Cause I mean startups is a competitive
world,

you know, like there's companies out there
that are locked in,

like right now, you know. And I used to
think about this back when

I started Laravel. I would absolutely not
be outworked

by anyone for sure. I knew that.

~ you know, back when I first started
Laravel in twenty eleven,

twenty twelve, I was like, okay,

well, they may be like smarter than me or
like more like advanced than me,

but like I'm working gonna work harder
than them.

And like in a on a long time scale,

it's just gonna compound and I'm gonna
beat them,

you know, just out of sheer like grit and
determination.

~ and that but that's that can be
challenging as the company gets larger,

it's remote, like

It's hard to maintain. So I try to,

you know, bring that to bear a little bit.

Yeah. Well, it's been really fun.

And I I I I love what you've described
here.

It absolutely makes sense of ~ you are a
self aware,

~ you know, you can read the room,

you understand people, and so as you bring
in a whole bunch of people

who especially like a lot of people coming
from the existing startup world,

and a lot of people come from the
existing,

you know, enterprise corporate world,

people who do things differently than you.

And there is a type of founder who would
come in and say,

Mm.

You all are going to

Co work my way. You are going to it's
shape to be like me.

You're at my company now. and then there's
a founder like you who says,

I want to learn from you and I understand
and respect what you're bringing

and all those things. ~ but in doing so,

you can lose yourself a little bit.

And that's ~ that's something I'm very
familiar with having done in

my personal life and in my professional
life.

and you hit a point where you're like,

wait a minute, you know, like I've
learned,

I've listened, and I've also squashed
myself and I want to.

Mm, yeah.

To to say that I not to say to disrespect
other people's opinions or

to say they were wrong, but you say like
there's also needs to be space for

me and I have been so good at deferring
and listening and considering others that

I forgot to be myself to. And so I'm I'm
like,

~ I love you doing this. This is so good
for my friend and it's so good

Yeah.

Yeah. I

for our community.

I feel that too. And I think it it really
I think what you said really hits home

for me too. And it it goes outside of
Laravel.

Like just I am a pr I as a person am very
like very like compromising

and willing to let go with the flow and
not really that hard to get along with,

you know, not really polarizing,

like in person, I mean like, you know,

just in my interactions with friends and
family and stuff.

~ but yeah, you can kind of like squash
yourself so much that like

Yeah.

It's actually a disservice to the to the
company and to the people you want

Yes.

to be working with. It feels like,

I'm kind of like I'll I'll be this person
that just sort of gives everyone space

to do their thing and not really bring any
particular flavor to bear.

Yeah. Yeah.

And I think that's, you know, that's not
good either.

And ~ you know, so that's what I'm trying
to correct for a little bit is like

be a little bit like crazier, be a little
bit more opinionated,

be like a little bit more genuinely me.

~

Yes.

Because I think that's just another that's
another,

you know, ~ ingredient in the recipe of
Laravel that is needed.

And without that, ~ it I think it is like
missing something.

Yeah, I couldn't agree more. I'm I'm very
happy for it.

~ I know we're just about at time,

so I did want to kind of end with just a
little bit more talking about you.

I get to know you as a friend,

and so I get to kind of like see what
you're doing in your life,

but I know everybody doesn't. What is
something you've done outside

of work recently that was really you know,

did you guys travel somewhere or have you
taken up hobbies,

or have you gone is there something you or
you and the family or you

and Abby have done recently that's just
like,

man, this was so great. I want everybody
should do this or I want to cheer about

it or anything.

man, you know, it's been a busy summer.

~ well, we did all go to Tokyo.

~ this is not exactly work related,

but I was there for Laravel Live Japan,

but we all went, we had to turn it into
like a family vacation.

It was very different actually than all of
the ~ most of the other vacations we've

Uh-huh.

done. Tokyo's massive, by the way.

Okay. Yeah.

It made like New York City feel small.

~ but it was f it was fun for the the kids
to like go to Asia in general.

Really?

Mm-hmm.

You know, I had been to India and and
other places,

but they had never been and just all hang
out in such a huge city that

is is quite different, you know,

than America. ~ and it that was a lot of
fun.

~ my son James, he's been fishing a ton,

so we've been chatting like fishing
strategies,

baits. He went to a fishing store in Tokyo
because they like make a special kind

Yeah.

I don't

of like bass baits, which I didn't know
that fishing was ~

Especially like bass fishing and stuff
like that was a big thing in Japan,

but apparently it is. ~ so we packed up
the suitcase with a bunch

Neither. Yeah.

of baits and brought them back to America.

That's awesome.

And do you fish? I feel like he's got like
an uncle he fishes right with,

right? Did I make that up? Uh-huh.

Yeah, I I can fish, but ~ he fishes a lot
honestly with his friends and then yeah,

it's with the friends. Okay, got it.

my my uncle

has a boat and takes him out. But ~ yeah,

Okay.

fishing has been a big thing for our
family for the last year or so.

~ yeah. But it yeah, it's been fun.

Yeah. Okay, and and

the kids, I I don't know I know we try not
to say exact ages,

but I feel like the kids are, you know,

late middle school, early high school ish
ages.

Yeah.

So you guys are staring down the barrel of
empty nesting.

Have you thought about that at all?

Yeah.

Is there any big changes coming in the
future?

Man, we have

Or like, we'll get to it when we get to
it.

I I can't say we've like cracked it on
what it's gonna look like,

Yeah. Yeah.

but definitely we've talked about it
because you know it's like coming

up relatively soon, say in the next like
seven or eight years.

we're kind of we're kind of there.

I don't know what it will look like for
us,

you know. Like, I grew up here in
Arkansas,

but my wife did not. ~ you know,

I my

My whole like personal situation looks so
much different than I thought

it would in general. And I think this is I
mean,

this goes back to trying to predict things
years in advance,

right? It's just so like impossible.

~ you know, my my parents have both passed
away.

Didn't expect that. ~ that makes,

you know, what is what where do we wanna
be,

you know, in the future? It's a lot more
like open ended now,

Yeah. Yeah.

like I, you know. ~ so we'll see.

~

Yeah. Okay.

~ I think I'll still be tinkering around
on tech stuff.

I'll probably still be tinkering around on
Laravel even at that point,

Yeah. Yeah.

hopefully if the company's doing well.

~ so you know, we'll see. But it I I think
it'll be fun,

~ because there's just so much it's so
open ended.

You know, there's we'll see where the kids
end up and then we'll we'll figure

it out.

Love it. ~ is there anything you want to
share or anything we didn't get

a chance to talk about today you'd want us
to cover?

well. Definitely stay tuned for Laracon.

There's gonna be a bunch of stuff.

We may have to do a follow up recap or
something.

Heck yeah, I love it.

But ~ yeah, there's gonna be a bunch of
stuff ~ coming out.

I'm really excited about it. I think ~ a
lot of it is stuff I've been excited about

for a long time. And Laracon is always
kind of this moment to like,

you know, take like a deep breath.

Finally, like all of the things are out
and we can talk about them

and there's no like secret projects we're
preparing behind the scenes,

Yeah.

or at least we get to start fresh doing
that again.

Yeah. Well I'll see if I can either steal
you away at Laracon or get

~ but no I couldn't be

you the week afterwards so we can do a
follow up.

Sorry interrupted.

Yeah. But

it'd be fun. I'm proud of the team,

you know. There's been a lot of like heavy
technical lifting and work that's gone

into a lot of this stuff. And ~ you know,

but that's what really makes the products,

you know, great is ~ tackling these sort
of hard problems that are very hard

to solve on your own with your own sort of
like small team or DevOps team.

These are pretty complicated spaces that
we're playing in now,

but that's what really gives the product
like value and ~ differentiates it,

you know, from what's out there.

So I'm excited.

I love it. Well, thank you so much for
your time,

especially and I know that you're super,

super busy this week. ~ I will see you
next week.

A lot of us will see you next week,

and then I'll I'll try and catch up with
you a little bit afterwards.

Awesome.

So it look forward to so we've got this
Taylor episode,

we've got a ~ episode coming after that,

and then after that it'll be one more
Taylor before we wrap the season and

we can do a full debrief of everything you
announced at the at the conference.

Nice. Yeah, it'd be fun. We can ~ talk
about everything.

All right. I love it.

Well, thank you for hanging out,

and for the rest of you, we'll see you all
next time.