Laravel Creator Spotlight

Alfred Nutile dives deep into the intersection of Laravel and the growing influence of Large Language Models (LLMs) in development. With over a decade of experience, including introducing Laravel at Pfizer, Alfred shares how LLMs are shaping the future of programming and content creation. We discuss how he leverages LLMs in his current work, his thoughts on the potential they unlock for developers, and what it means for the evolution of the Laravel ecosystem.

Show Links

Alfred Nutile - https://bit.ly/m/alnutile
LaraLamma - https://docs.larallama.io
PHP and LLM's the book - https://leanpub.com/php_and_llms
PHP and LLM's the Discord Channel - https://discord.com/invite/CKj7kVQDVU
PHP and LLM's the NewsLetter https://sundance-solutions.mailcoach.app/php-and-llms

Creators & Guests

Host

Eric L. Barnes

Creator of Laravel News

Guest

Alfred Nutile

PHP and LLMs author - get a sample at https://t.co/bEmA1cWr06 or purchase at https://t.co/Wl5ia9ZCb0

What is Laravel Creator Spotlight?

Interviews with creators making cool things in Laravel

Welcome back to the

Laravel Community Creator Series.

Today we are joined by Alfred Nutile,

and he has, you know, he's been a pillar

of the community forever.

I think I've known about him for, man, 10

years now? Does that sound about right?

But, you know, Alfred, welcome to the

show, and if you don't mind, sort of

introduce yourself to

those that might not know you.

Right, no, thank you. Yeah, it's been a

while. I kind of realize I'm

not in the background a lot,

especially I was hidden away at Pfizer

for a while. I introduced Laravel

there eons ago, 10 or

more years ago, 4.2.

And I worked there for like 11 years as a

contractor, kind of hiding away.

But for the past three years, I've been

on my own and just producing more content

for the public

because of being on my own.

It's a lot easier just

to share things. So, yeah.

That's awesome. Yes, yes. I've, man, it's

funny how like, you know,

the community's been, you know,

it's just so big from all these years and

like you don't get to meet everybody just

because it's, you know, now it's so

global and worldwide.

So it's awesome, you know, for you to

come on board and do this with us.

That's awesome. Yeah. Thanks. So you have

a project which is pretty interesting.

It's called Lara Llama.

Did I pronounce that right? Kind of like

the animal? Yeah, yeah.

It was called Lain Chain before that or

Lara Chain, sorry. But people got to

confuse with like

Bitcoin and that type of stuff.

So and I made it after Lain Chain, the

Python version, but now it's just called

Lara Llama just because again, naming's

hard and I just tried to rename it.

That's good, though. I mean, that seems

memorable, you know. It's

like, oh, that's simple.

And, you know, it stands out because it's

not something you hear very often. So

what exactly is the project?

So and that's the thing. So two years

ago, Lain, so basically it all started

out of being scared out of my socks one

night learning about Lain Chain.

Lain Chain is the Python project that

builds automations and workflows and

makes it really easy for developers to

say, hey, I want to connect all these

things together with LLMs and workflows

and all this stuff to then make things

happen, projects and, you know, agents

and stuff that can really automate stuff.

And I'll talk about that later because

with your workflows, we could talk about

that. And I was like, wow, what's going

to happen to me, a PHP Laravel developer?

Do I have to start learning Python?

Where do we fit into this next era of web

technology? So then like around that

time, I started making Lara Chain to just

prove to myself we still have a place.

And when you come down to it, one of the

biggest things with open AI is they made

an API that was really easy to use. And

we, Laravel is really,

really good with APIs.

And so when it comes to building web

stacks that can have scheduling,

batching, automations, we still have a

foot in the game or whatever they say,

because we can do all that and use these

APIs and get all the benefits of this AI

or more importantly, or more

specifically LLM integrations.

So that's where Lara Llama is at now at its

core, it is a rag system, but where it

helps to retrieve data from a vector

database augmented with the LLM and

generate more data

from that or more output.

But it's more than that, the goal was

actually to go beyond making it for

developers, but making it for business

owners or not no code level people as

well who can log in, click, click, click,

and do things to automate like checking

your email boxes or scraping websites to

gather data about their clients and then

generate more content.

So it really started to grow into more

than a rag system because I'm seeing

workflows and automations around that.

And one of the most recent ones I'll talk

about after is how it can be a personal

assistant around projects

and planning around that.

So that's the gist of where it came from

and what it's about.

That's awesome. Yeah, it's the whole

like, you know, for me sort of very new

to that side of the tech stack, you know,

tell this the future

of LLMs and everything.

Like, say you had to explain LLM to

somebody that's not even big into coding.

How would you even

describe that, I guess?

I know a lot of people are saying it's

auto correct on steroids. I think I think

a good way to look at it is, you know,

it's just a way for us to feed a prompt

to or a question with some information

and get results and it basically, you

know, people are fearful about

hallucinations, but it doesn't have to be

a problem once you get

better at these tools.

So an LLM is just an API or system or we

can use to basically hand it a question

and get results. So, for example, I have

a lot of I'll use it. Here's a nice one.

We send recipes to Lera Lama via the

email address it makes for you.

It takes all those recipes and at the end

of the week, it sends a summary of the

recipes and what we like, you know, it

then summarizes them by title, by cook

time or whatever we want, because we tell

the LLM in the prompt, you know, at the end of the week, send us an email.

Summarize all the new recipes this way

and then it produces that output that

way. So does that

answer your question there?

It does. It does. That's pretty good.

That was a good explanation, I think. So

I think my next question just about the

project is you can actually use sort of

any back end, right?

As far as like, yeah, or called or. Yeah,

that's a good question.

There's so more, right?

Oh, yeah, yeah. And it's endless because

again, we're like once this became an API

thing, it's endless and we shouldn't lock

ourselves down to a API

like open AI, especially.

So the goal here is two things. One is by

being agnostic to the LM where we can

talk to any API and get the results, just

like we do with storage or cash or

anything in Laravel.

We talked to an LM. We get the results in

we format them like the same for any LM

you're talking to any API.

And so at that point, it doesn't matter.

You can just keep spitting out these HTTP

driven drivers in Grok, Claude, Olama,

which that's why I really want to talk

about an open AI are the ones

that I have working right now.

But then you get into the more important

thing coming up down the road is like

within two years, like recently, PCs have

released chips that are good with AI.

So you can run Olama. Have you heard of

Olama at all? I have not known so Olama.

Olama. Yeah, Olama is like the engine X

of LLMs. Okay, you run a llama on your

machine, just like engine

X download, click running.

But then you pull in a model. So I'm

going to pull in the model, llama 3.1. So

now you can then run that model on your

machine and ask it questions.

It makes an API. So just like engine X,

it's a proxy. That proxy has an API, you

talk to that API, it happens

to match opening AI exactly.

So now you could just point your driver

there. Right. So now you have a local LLM

that you can talk to, more importantly,

for a lot of people, a, it costs nothing

besides your machine and B, you don't

have fear of like your data leaking out,

because you're running a local LLM.

So that's another win here is that when

you're working as a developer locally, or

if even for privacy reasons, you can use

that local API. So it's another key thing

there. So that's awesome. Yeah. Yeah, it

is amazing. Yeah. Yeah.

I was gonna say, yeah, it's, well, let's

just, while we're talking about this,

let's talk about some

use cases. Yeah. Right.

You know, for me as a sort of a content

creator, the first thing that comes to

mind is like, article summaries for

sharing on socials. Yeah. That's sort of

the first one I think about, but there's

like so many use cases you can use this

for. Right. I mean, it's, it's sort of

endless, whatever you, you

can dream up, you can do.

It's good. It's yeah. And it's so let's

talk about use cases. And then along the

line, let's keep in our minds that, hey,

chat, you can do that. I can go to their

website, paste this in, and it can do it.

So we'll always keep

that in mind as we talk.

And then another thing to keep in mind is

like, no matter what it can do today,

we're always going to be ahead of it,

because we're developers, we're creating,

we're thinking ahead of things. So we'll

always be able to offer our customers and

ourselves an edge because

we're always ahead of the case.

And I've proven this with myself for

years, at least two years now with

opening AI, I'm always way ahead of them

on some ideas and abilities. So back to

your point. You could use like, like, let

me specifically talk about mine. And then

you'll see your overlap.

I made a project in Leralama. That

project is basically me with a prompt

saying, hey, help me market this book.

What are some key things I can do? Build

me a plan. And then the project says,

okay, here's your plan.

Here's all the tasks you have to do. And

when they're due, I'll remind you every

morning what's going on this week. Right.

And then in the background, I can have a

bunch of other stuff happen because

Leralama has these plugins or these ways

to send data to the system, which is

another thing to keep in mind is all day

long, we're collecting

data or we like things.

And if you can send them somewhere

consistently and then have that be your

repository, then that's a big win. Where

it's opening AI or chat, GPT and all

these web interfaces, it's kind of hard

to just keep it up to date and keep your

data in there. So

another short term win there.

So now all week long, I can send it via

email, a link to a news article I like.

So 10 articles later, it's collecting all

this stuff when it gets the email, it

goes, oh, there's a link in here. I'm

going to go his prompt says go get the

website from that link.

And I'll go get it goes, it gets the

website downloads it and puts it into the

system, right? So now by the end of the

week, I have a collection of emails that

are now websites I wanted

to talk about or think about.

I have a system that could help summarize

them for me. I could potentially have the

system plugged in the LinkedIn or Twitter

to help post for me consistently,

especially once I build trust with it.

And that's another thing I talked about

in another article of like, you can automate a lot of things, but we can't do that.

We can automate a lot of things, but we

have to build that trust meter in the

specific task. So for you, it's it kind

of overlaps what you're doing, right? You

can start collecting information in one

place, you could start helping it, it can

even know your voice and your tone.

So there's an area in Laryllama where you

can say, this is how I speak, this is how

I write, I mean, right? So then when it's

repeating things or building things for

you, you can say use

my voice or use my tone.

And when the LM is building that response

or building that summary, it can then use

your tone to help do that.

So that's that's pretty sweet. Because

one thing just using like chat GPT is the

when I write prompts, I try to be really

specific, but it's like it will still

come out weird a lot of times.

Yeah, so as an example, so like when we

publish a new post on the on the Laryll

news website, I'm just using Zapier right

now and it pulls down the article tags.

Yeah, it runs it through chat GPT and

supposed to summarize it into a tweet.

And then it pushes out to like threads

and LinkedIn or whatever.

But a lot of times it'll come out like,

like in a first person, you know, like

I'm writing about something Larylla

Incorporated did. And it'll be like, look

what we did. And it's like, but that's

not that's not even how I wrote the

prompt. Like it should never do that. But

you know, look what we made in this

package. But I've still been struggling

with the trust issue on

the way I've been doing it.

Yeah, in in one thing to mention there is

prompt engineering seems like a silly

concept. But I think you'll see you have

to get better at it. Right. So just sit

there and hammer away in the in the chat

GPT UI or Claude, I like Claude better.

Throw your prompt at it, get your

responses, get them right, keep messing

with the prompt, ask it to help you fix

the prompt to get the right answer

because you're what you're asking isn't

hard, impossible. I mean, it's just, you

need to keep refining your

prompt to help with that.

And you know, the concept of multi shot

versus one shot, like you can give it

multiple examples of what you want

multiple examples of what you don't want

so that it can keep learning and not

learning is a bad word. It can keep using

the context of your prompt to

not do the wrong thing again.

So yeah, awesome. Yeah, yeah.

Not to sort of derail too much. But as

far as like deploying like say I want to

run this on my own on my side, I can just

deploy it as a normal level app or is

there like a lot of stuff involved like

I'm thinking like a

forge or something like that.

Yeah, yeah, forge.

It's really that easy. So, you know,

ideally it would be a bunch of easy like

filament ideally it would be like

filament but it's not like, listen, my

goal is to inspire

people to think about this.

Right, we have it, we have, look at I was

pulling up some numbers about lane chain

there's 72,000 stars and

GitHub versus level 78,000.

This is a popular thing that no one's

talking about in our community very few

people that I meet even know what lane

chain is. Okay, yeah, I've never heard of

it. Right. So my point is, there's

something going on here it's more than

hype, it's a new way of thinking about

tools just like WordPress was a CMS.

We have a new era of like I call it a CRS

a content retrieval system but it's more

than that it's automation it's workflows.

And so, ideally, it would be like

filament where you can plug it in your

level package and go right right now it's

a, it is like WordPress where it's like

the whole thing or nothing, and that's

not ideal, whatever.

But honestly forage deploy use a Postgres

database on forage in everything should

just work because it's that simple just

Laravel Postgres with its vector ability

which is awesome horizon to run q jobs

in, and that's pretty much

it in web sockets reverb or

to get a little bit of the dynamic UI.

But then what you get after that is you

get the the awesomeness of Laravel with

the scheduling where you

can have tasks waiting to run.

You can tell it to go check an email box

so that's another thing.

I have an email box so you can make a,

you know, like an info at your email

domain and then you can say

okay go get emails from there.

And so every time you make a new email

box it will just put the email plus 12345

at that domain so you can just keep

creating on the fly more and more places

to send different types of emails.

These are my emails about news articles

these are my emails about things I want

to look into, blah blah blah so before

long you can use these

different techniques.

Or how about this one you have five

projects and they're all in GitHub and

every time you commit the webhook sends

back to Lara llama the results so now at

the end of the week, you can sum up your

change log for five projects

because it's all right there.

So you got webhooks you get web scraping

with browser shots you

can set up browser shot.

And you have web search with a brave API

right now so you get all these things

like you said by just forge deploy

postgres which forge will do for you and

it should be that easy honestly.

Yeah, that's that's really good. That's

what I like to hear is the easy side of

it. That's the goal in it hasn't so

there's the concept of output is there as

well where you can say hey make an API

API around this collection of data so now

you have a collection of news articles,

or more importantly.

I was talking to someone like, well, I'll

just say like there's a company I know

that has articles online. And they're

like, how do we make this whatever like

like so it's using this modern technology

but you could be

sending all that stuff here.

And then you could say, make this an API

click a button, and now you have an API.

So now on your website somewhere else you

have a chat widget that talks when the

user asked a question that actually is

asking it from your API and layer llama

and gets the results so you can actually write away.

Plug in this LLM API to any website you

want, but using your data that makes it

chattable. So yeah, you can go in a lot

of directions there

with with layer llama.

And again, this is all to inspire ideas

if layer llama has gone in five years. I

don't care. It's to get us thinking as a

community. What can we be doing in I

think what we need to be doing this state

may be relevant relevant relevant to

relative in this next era, you know.

Yeah, for sure. I mean, that's, I mean,

that's all you keep hearing about

probably the last what five years is, you

know, is open AI and lms and everything.

Yeah, I'm engineering so it's I feel like

it's going to be it's one of those things

that that's definitely

going to be here to stay.

Yeah, yeah, yeah, I have a question. This

is totally off the wall but you know, as

you're describing all this so one feature

of the Louisville news site is we have a community link section.

And it's, it's sort of hard to keep up

with because you know I'm only one person

and then what I normally do is every

Sunday is when I go in and

approve them, it would be nice.

I don't even know if this possible but it

could could it's good. Yeah, I am or not

spam. No, that's beautiful. Like, 100%

like you could say hey, go look at these

URLs go parse them so layer llama and again, you can grab the code.

Just use it so layer llama will take your

URL if you use so so lms have a tool

ability and the tool says the LM says oh

he wants a URL turned into a website so

sometimes I'm going to use my tool, you

know, URL to website and

it just knows it right.

So then it would get that data and your

prompt will say then verify if this is

spam or not and then the prompt.

The results would be like yes or no like

whatever you want. Right. So you could in

one prompt. Say, check these URLs for me.

If it's spam return false

if it's not spam return true.

And then even you know do the next thing

if it's okay then your next tool would be

then go set it to approved in my database

because it could query your database but not magically.

Think about tools dependency injection.

Right. So when it says

use the get URL tool.

There else dependency injection system

gets past that line that then renders

that class and it's past the arguments or

the information and then do what it needs

to do and then return whatever it needs to return. So there's no magic here. It's just it knowing what tools you have what steps to take and what to what what to do with the results of that tool. Right. So yeah, I think that's awesome.

You could do that and that's a brilliant.

These are the things I want people

thinking about though you just like it's

not to take away your job or many jobs

but it's to assist us in getting this

stuff done that we don't want to do or grunt worker. You know we can be doing something else. Yeah, we mean the way you describe it could even you know it could see if it's spam.

But then secondly it could be like can

you make sure this article is actually

about Laravel? Yeah, 100% by that. Yes, I got I got to have to research this more because that would be awesome to have that more automated. And I think what you saw there hopefully one thing I am writing the book in in will talk about in a moment, but basically we're empowering you to do that.

I got to have to research this more

because that would be awesome to have

that more automated. And I think what you

saw there hopefully I one thing I am

writing the book in in will talk about

that in a moment, but basically we're

empowering users to like like not only is

business owners like yourself, but when

you build an application and you could

say to a user number all those things you

had to do to query the database or to

filter or do this. No,

just type in that prompt.

Find me people who live in California who

went to the school and it does it all for

you. You're empowering users to get their

jobs done without horrible interfaces or

tons of steps. You know, so it's that

prompting ability is a

key thing here. Yeah, yeah.

Yeah, well, actually, this might lead

into your book, but as far as tips for

good prompt writing the best is it best

just to sort of just trial and error and

ask it. Like you said,

just ask it to help you.

Um, yeah, I mean, I would. What I there's

a lot of bad information out there where

everything gets more complicated than it

needs to be. If you stick to a particular

structure, you'll do pretty well.

And then there's the deep learning dot AI

site where it's written by or run by

Andrew and G. I think I

forgot to say his last name.

And there's a lot of good trainings

there. Even one of the ladies who works

at open AI had a good prompting training,

but in the end, basically start off with

the role in brackets and say, hey, you're

my assistant who's helping me parse this

information to find good or bad links.

And this is, you know, these links are in

explain what it's about. Right. And then

your task is the next bracketed area.

Your task is to find the links that are

good and ignore the ones that are not in

at the end, you're going to return to me

an array of results.

And then the format is the last bracket

second to last where you're like, do this

format. Like we just said, it's an array

in the object with the URL, the title,

and maybe a TL DR or all the content.

So it's role task format. And then the

last one is just context

where you're handing it.

If you're in your case, you'd be handing

it the links, right? Where you're like,

here's all the links I want you to look

at. So it's that particular format that

is a good starting point to then just

start cranking away.

And then, you know, you have examples

like don't do this, do this, you know, I

didn't like this. And then you can learn

from those. Like one thing I, that's kind

of interesting is when you

use chat activity or cloud.

And you're typing in that prompt. And you

say check if you go do this, there's a

system level prompt that's probably this

big that's telling chat activity what not

to do what to do how to answer you how to

treat you how to think like, it's not

just an LLM getting your question.

It's an LLM who has a prompt around your

prompt. Okay, so, so you got to remember,

you know, it is a big deal prompting and

you can give it a lot of information.

So it can be helpful to give it a sense

of what it's doing, why it's doing it and

the task to be done. Does that make

sense? It does. It does. It's, it's, it's

funny. It almost reminds me of a back

when I was in school, one of our

projects, this was like middle school was

like we had to write down how to tell

somebody to make a peanut butter jelly

sandwich if they, oh, yeah, and they can

only follow the steps

that you tell them to take.

Just I just remember how hard that was.

And then like, yeah, so specific, like

you pick up the knife, you dip it in the

jar. So it has a lot of, to me, it has a

lot of feel of that to this, you know,

yeah, don't give up. No, it's a good way

to put it. So don't give up because you

told it to do something you thought was

obvious. But keep going and keep it

giving it more details.

There was a funny video on what you just

said. And the guy had his kids doing just

that. And he followed the directions. And

it was just silly. Like, you're like,

yeah, okay, I see. You know, it was a

good one. You know, that's awesome. So

you mentioned, you mentioned the book. So

you're writing a book on PHP and LLMs.

Yeah, tell us a little bit about that.

And like, you know, how

that came about, like, right.

Right. So basically, it's been a couple

years now. And I use it every day, not

just to help get my work done, but with

clients. I have a lot of moments where it

just it's a tool that solves a lot of

problems for me or gets me, you know, to

solve a problem, because I didn't

understand like how to do GIS in a

database and actually know it has me

doing everything I need to do. And I was

able to pull it off or format emails,

data from emails that were coming from

random people to then generate JSON

consistently without having to do that.

And I was able to do that consistently

without having to write all the text

parsing I was trying to write. So I just

wanted to start sharing those as

practical solutions in this book, almost

like a recipe book. But at the same time,

help the developer understand how this

works with prompts, how this works with

drivers and APIs and how we can build or

work in this way that we don't it doesn't

matter what API reason because it's, you

know, and it will continue to do that

where it shows all these

examples, explains it in

hopefully at the end, the person can just

start to see real like day to day

examples, I can use it in for work. And

it just it's, it's really interesting

stuff because, you know, it's a lot of

stuff is changing. And so I want to it's

kind of the foundation. So you can keep

ahead of things and not wait for things.

Like, if you're waiting for open AI to do

something or our PHP, open AI driver,

you're gonna be waiting. And you

shouldn't be because these are just API's

and code just make stuff happen. So

hopefully, we'll help people get

ahead and just understand the root of

things. So that's awesome. So, so the

book, it's, it's still in in development,

right? Yeah, it's still working. Not not

officially launched yet. Yeah, it's about

five chapters in, it's there to buy. It's

there to see some sample chapters. And

there's a newsletter that you can then

keep up to date with the articles and

stuff going on. And I will I continue to

release some of the chapters or articles

about the chapter to just show

you some information while I go. So

that's awesome. Yes. And, and so where

are you? Where are you publishing the

book? Are you using right? Yeah, I went

with lean pub for no good reason. I'm old

school. Yeah, I saw I could think of. So

yeah, I don't know if anyone uses them

anymore. So yeah, lean pub, PHP, LLMs

should be able to find the book pretty

easily. Or we can share a link after.

Yeah, yeah, it will definitely have links

of all this in the

show notes when we when we

push it up. And then I guess, you know,

as far as the the book, the PHP, LLMs, is

it? Yeah. I mean, if you know, Laravel,

you know, PHP. So it doesn't it just

raising the title PHP gives your

audience, right? Then yeah, this is it.

And I didn't want to deal with any legal

stuff. But it's so Laravel. It's insane.

I'm a Laravel. It's what I know. You

know, so but it is, you know, at the end,

I want all PHP developers to

kind of be aware of, like, the

Python community. Let's embrace this like

the Python community did, let's make

tools around these APIs and, and so

forth. But yeah, it's a good point. It is

very Laravel centric. That's awesome. Are

you thinking of sort of like creating

like a community around all this as and

being like the the thought leader of the

LLM PHP LLM? I don't know if I have the

control to do all that. Like, I think

this stuff just happens

organically who you know,

who you know, where you are in the

community. For me, I'm just going to keep

I have clients, I'm working on things for

people, things come up like, you know,

it's just amazing the stuff I've been

solving for people lately that I couldn't

have done before. And it's just fun.

Because when you were like, Oh, crap, we

can do this, you're like, this is

awesome. Like, even one client, like they

have this list of data, they have to get

in the client in the customer with them

export it because it's, they're hoarding

your data. You know, she can take

screenshots and it gets translated

into the system. Right? Just from a

screenshot. And it's like, yeah, this

saved our butts. You don't have to hand

enter this. This was awesome. So there's

all these little moments that come along

where problems are getting solved that I

couldn't solve before. So I'll keep doing

that. I'll keep blogging. I'll keep

putting stuff on YouTube. And I'll have a

video course eventually after the book to

just help people go deeper, hopefully, or

just continue to hopefully learn this if

they don't like reading, they rather

learn through videos, some people like

that better. So yeah,

the community thing.

I hope, but it's a tricky one, right?

Like, it's there for the open sources

there to be a community. And hopefully

people can drive it to that next level. I

don't know how they did it with filament.

I don't know if it was just good stuff at

the start, or like, but that's a great

example of a success in my book, that I

would like to strive for with it. Yeah.

Yeah, yeah, actually, I don't know. I'm

feeling that, you know, I know, Laervel

itself, you know, I was there in the

very, very early days, and it was

basically, you know, our C, and it was

just like, that's how we communicate and

talk to everybody. And then sort of. And

then, then we then they switched to

something else. I can't

maybe it was so like, or not.

Telegram.

Telegram, yes, Telegram. And then from

there, it went to Discord, I guess. But

yeah, it's, it's sort of, you know, I

think just, you know, as a project owner,

just having some place to go and be like,

where can I go ask a question when I

don't understand what

is really beneficial. But

now it's a good point. I'll keep that in

mind. I was using GitHub's discussion

area, but it's not I don't think it's the

same because it's not real

time. So it's a good point. Yeah.

So,

you know, speaking of on that, this is

totally on a tangent. Now, the one thing

I hate about discord and all these other

things is you can never search like the

search is never good. Like, you know, in

the old days when everybody used like old

style forums, you could just go to Google

and type in like

whatever and it was there.

Because it was all, you know, indexable

and searchable. And it was like, ah,

well, this is the answer. But I guess, in

theory, we're actually in this

transitionary phase of searching right

now, you know, because you can still use

the Google and you still get the, you

know, whatever. But now it's like,

everybody's going to chat GPT or going to

all these things and searching instead of

actually going to Google, you know, if

it's something that they can explain, and

you're getting better results.

Yeah, I mean, they're going to release

something soon about that. Yeah, I talk

about tangent. Yeah, I just discourse and

I always get the mix up as a discord or

discourse because there's that one where

the stack overflow guys were released,

which was a great forum software. But

then there's the one

everybody uses for everything else.

And I always click the video button. So

in the moment, I'm like, panicking

because I'm like, I just started to call

with this person. I don't know.

Yes, yes, I think yes, discord, C O R D

is the one with the call button and then

discourse like the form.

But I'm going to follow up later, because

I think you're spot on. I need a more

dynamic place if I want to build

community where people can come in and

ask any question. Yeah, that's a good

point. Like with native PHP, I joined

that one. So I get to do one,

too. Yeah, it's a great idea.

I love that. And now the book when when

you're expecting, I guess it's for sale

now, and then you just release new

chapters, you just if you purchase the

book, you get the chapters as you write

them. And as they come out,

yeah, they come out every week, one or

two on Monday. And then if at the rate

I'm going, it will be done by early

November. And if I remember, right, and

but along the way, I've been adding

chapters, I think you asked it recently,

you said, Well, why why LMS? I'm like, I

don't even have a chapter on that. I just

assumed people know why. So I built out

that chapter. And so that's a good

example. Sometimes I think things are

kind of come in as I as I do this stuff.

You know, so but by early November, it

all should be done. Yeah.

Awesome. And, and then this, I don't know

if your website says this or not. But

let's just say I want to hire you to be

an LAM for something is that is that is

that a possibility? Are you? Yeah, of

course, of course, you do custom stuff.

I always do. That's all I do. So I full

time, I'm a freelancer who just does

custom stuff for for customers. I'll give

you my bit.ly link at the end, because it

goes to all these, these things,

including how to get me a calendar

invite. So you can just talk to me

sometime. But yeah, if you ever need

consulting, or more importantly, build

something fun, I can help out. I mean,

some of the stuff I don't want to talk

about, because it's kind of NDA, but just

the stuff you can do, it's just, it's

just like with you and you're automating

your workflows, like,

you'll be amazed at what you do. It's

cool you're using Zapier, I never think

about them, because I'm too busy trying

to make that thing. But I forget how easy

it can be if I would just stop being so

stubborn, you know, well, I mean, it has

its benefits. But it's also very

expensive to run. It's it's

not good. Good point. Yeah.

Because now they charge you by the like, you get so many tasks, tasks run by month or something in the way it works is every link we approve goes out to, you know, for social services or, you know, five, it goes out
everywhere. And then, and then we publish posts that goes out everywhere. So it's like, we're running tons of tasks. And it's all like, yeah, probably 1000s of months, I guess, is pushing all that stuff out. But well, we should look at

we'll look at lara llama together, I'll

show you my hosted versions, my training

version, and then we can just see what

can I do? What could it do? And then

maybe we can play around. Because in the

end, I love these real use cases, right?

This is how products succeed is they have

real use. I think that's why Laravel

succeeded. It wasn't trying to solve

imaginary problems. It was a SAS focused,

like product from the start, you know, it

was solving real problems that, you know,

paid off. So like with you, if you want

to talk about it, we can just, you know,

I'll show you around.

And show you it hosted and show you all

that. And then you can see what can you

do now? And what could it do easily? If

you just think about

it? Yeah, so it'd be fun.

I love that. Yeah, that's a that's a

wonderful idea. Yeah, yeah. Yeah, we'll

definitely have to get that scheduled.

Um, so sort of to wrap up the call.

Anything else I missed that you know,

might be you'd like the

community to know or anything?

Right. I mean, I will start a place to

talk more about this. But I guess the

only thing I just keep saying or

hopefully showed in this or explain this

like it like it, there is hype and the

hypes gonna fade because hype fades, I

think it's what it has to be hyped to

fade. So, but I think in the end, this is

a real this is a new tool that we can use

today to do some cool stuff for our

clients and for ourselves.

Like, like a total home project I can do

now. There's so many of them. But like,

like, just a side note, like just how

cool this is, like now you're a

developer, you're thinking you're

tinkering, you want to do things. Like no

one ever in my household says, hey, we

ran out of oatmeal order some more empty

box, they don't even take the box down.

It's empty. But just say they took the

box down, threw it away. And then

tomorrow, they're like, hey, there's no

more. I'm like, yeah, you ate it all. No

one told me to buy more. Right. So, so

like, what I did was I put on a

I haven't finished this yet. This is just

an idea. I have this top row of all the

backup food. This is our when this is

gone, I have to order more. And now you

take a stupid little camera on a

Raspberry Pi, and you point to that and

take a photo. Or every time motion

happens, and the LM could say, Oh,

oatmeal is missing. I'm going to now

order more. Because it happens to be

right next to Alexa. And it can say,

Alexa, order more oatmeal, like, you

know, the fun stuff we can do at this

next level for for just

messing around. It just it's kind

of cool. So I would just say, like, just

see this stuff is like, it can help do

some really cool stuff we couldn't

imagine before. You know, yeah, that's

such an awesome use case. Because every,

every week when we go to order groceries,

always like, what do you want? I'm like,

I don't know, just normal stuff. And then

she's like, Well, what is that? Like, I

don't know. Like, look, I now made this

project, one of mine's a meal assistant.

So it's a meal planner, we send our

recipes, we tell it we like, and every

week, it can build up a

plan. It can send us a memo,

email saying this is your shopping list

for this week. And then every week, it

does another plan, hopefully not

repeating everything. Because that's what

I always do. I'm like, let's have this we

had that last week. Okay, like, so it's

just fun stuff. I know there's so many

More episodes

Chapters

Creators & Guests

What is Laravel Creator Spotlight?