the true cost of AI coding. Yes, the

very very true cost of AI coding. You

may have seen this video floating around

for this last week in which Scott right

here discusses kind of the cost or the

negative and the positive effects of

what AI has done to the software

development arena. I think a lot of

people have been feeling the same thing

lately. It kind of has been showing up

all over the place. And I think that's

why this video I'm done coding with AI

just went so far so fast because so many

people recognize a truth that is kind of

growing which is that those of us who

are actually really good at coding. A

lot of us have had probably the hardest

transition in the AI coding world

because the things that I loved was

sitting down and typing. I didn't have

to go look up the difference between

slice and splice and JavaScript. I

didn't have to go look up documentations

for how to write things in Go or in Odin

or in Zigg or any of the languages that

I got good at cuz I could just use them.

It was a flow state. I just sat down and

the ideas in my head came alive and my

fingers were the translation layer. I

was I am a wizarded vim. Okay, I'm a

wizard, Harry. But it's impossible to

ignore the benefits of AI. Yes, you can

most certainly you can ignore them.

There's a lot of people that will be

very successful ignoring the cost of AI.

There's people that were very successful

writing code in Notepad. That doesn't

mean the average person's going to do

that. And that's why I think just so

many of us feel roughly the same thing

that AI is here. There's obviously a lot

of benefits to it. There's also just a

lot of negative things. And and the

negative things, they don't have to be

bad code. Yeah, sure there's some bad

code, but I mean, bad code's always been

around. It's a matter of fixing things

and stating it the right way. But that's

never really the problem, right? Like

the problem wasn't the bad code. The

problem was I liked just sitting down

and programming. But I also recognized

that that's a dying activity. That's an

activity that few people will still

enjoy. And slowly it will be relegated

to those who just do it for the sake of

doing it. Kind of like riding horses.

Fun side story. Uh I had a cow escape.

And the only way to get the cow back was

to actually have an expert horse rider

who did come in and rope the cow and get

it back onto my property. So, just to

throw it out there, it turns out there's

still people who make money riding

horses to this day. And so, this idea

that, you know, like horse riding

actually is just irrelevant completely

in a society full of cars actually,

well, it's not act it's technically not

true. All right. Anyway, sorry, back

back on topic here. Uh, but either way,

I wanted to talk about this video, the

true cost of AI coding, and more

specifically this fifth chapter. What

separates energizing versus unhealthy

use of AI? Now, of course, before we

discuss this, before we jump into the

meat and the potatoes, a quick thank you

to the sponsor, Vibe Coders, Silence.

Someone who's worked in enterprises

speaking. You know, simple ooth signin's

not that bad to program for real

enterprise applications that need single

sign on pass keys, user management,

roles and permission, enterprise

readiness and passwords and recovery.

uh that's a little bit more difficult

and that is where work OS comes in. Work

OS is trusted by a lot of the top

companies open AAI, Anthropic, Cursor

and a whole lot more. But not only that,

it's free for your first million users.

So that means for you that is 9999,999

free seats available. Pretty much no

matter what backend you have chosen,

there is an SDK available for you. So go

check out WorkOS at works.com.

Links in the description. All right, so

we're going to watch chapter five and

I'm going to yap about it. Yeah, we're

going to do an old-fashioned good

oldfashioned react here, people. So, I

also really wanted to know like what

separates energizing AI coding use from

one that's diminishing of your spirit or

energy or any of that stuff.

>> By the way, uh even before

AI and all this, there was energizing

and and kind of like diminishing

enthusiasm coding tasks. I think we all

understand this. There was this article

a long a long long long time ago that

effectively made this statement where

the most senior engineers can do the

most junior tasks because we are able to

kind of plow through a lot of the

monotony that a junior might struggle

with that we can actually stay t you

know stay on task stay focused and that

was like a huge level up for me in all

of my career was being able to be like

no I know how to say no I know how to do

the mundane I know how to take apart the

hard problems and just work through the

slog and that was a lot of programming

like most of programming isn't de you

know creating a compiler from first

principles a lot of it is just put pipe

A into pipe B make it happen even the

game I am working on at this point it's

like coming up with new card ideas and

the world's easiest programming like all

the programming to the game is just UI

and how the game operates there's

there's nothing left in it that's

actually difficult in other words It's

kind of monotonous at this point. It's

just a it's it's purely a gated idea

problem. I have to come up with ideas,

play it a whole bunch to see if it's

actually fun. Can I create lightning in

a bottle or not? If not, reject the

idea. Aaron Francis is the VP of

marketing and community at Larave. By

the way, Aaron Francis, good guy. Big

big fan. And he's a developer and

product builder. He's been building

things for a very long time. He reached

out to tell me he doesn't get any

negative mental health side effects from

AI coding, which is exactly why I wanted

to talk to him. By the way, that seems I

I think like anybody hearing that that

kind of feels crazy, right? Like I'm not

speaking of mental health side effects.

Look, cursor agent finished. Okay, guys.

I've been running this thing where I've

been creating this this uh testing

framework to test my game running. So, I

always have a bunch of agents playing,

searching for bugs and stuff, okay? So,

I get a lot of alerts. All right. I've

been getting a lot of alerts lately. I

definitely went through the psychosis

phase for sure. Um, and what that looked

like for me was just uh almost like

infinite possibility. And that part's a

little bit scary because you're like, I

could literally do anything. I think

that's is is extremely insightful. Let

me let me give let me give you a reason

why. Before when you had an idea, right,

you'd actually have to sit down and

you'd have to build out a little bit of

a prototype. By the way, pretty decent

square right there. Hold on. And we got

to actually do the square test. Square

test, boys. Did I actually hit an

allnatural square? I don't think I think

it's a little tall compared to wide. Uh

right now, what are we sitting at? We

are sitting at a 152x 15. Oh, yeah. 152

by Oh, wait a second. A 151x 154. Okay,

I think that was close. Okay, that was a

three pixel off perfect square. That

pretty good. That was pretty good. See,

back in the day, what ended up happening

is that you had an idea and the idea was

extremely not fleshed out, right? you

going, "Oh, I think this would actually

be really easy to be able to do." So,

you spent a little bit of time coming up

with an idea. And really, this is like a

backbrain idea. It's not something

you're actively engaged in. It's

something you're seeing every single

day. And all a sudden, you're like, "Oh,

I have an idea. This could actually be

really, really good." And then you spend

a lot of time coding, right? We're

talking about like 10x the amount of

time, 100x the amount of time you took

to come up with the idea versus actually

just implementing the very basic

version. And while you're implementing

it, you think through all these

different design ideas. And that way it

gives you so much like active time spent

thinking about an idea while you are

kind of like massaging the code and

using it. And so you're really coming up

with this like really tactile way you're

coming up with feedback. It's not just

simply you using the thing. It's also

you actually programming it and thinking

through its exact problem and going well

wait a second. If I shape the data this

way I can look at it this way which the

UI or the interaction can look like

this. and you kind of come up with this

like really cohesive kind of human

experience within whatever you're

building that actually makes it

compelling to use. And so then

theoretically the time you actually

spend coming up with the idea is larger

than the time you spend coding cuz all

the time you spend coding is also

feeding back into this idea design. By

the way, really bad square right there.

I I know you saw that. And then these

ideas while you come up with them, you

start integrating them back and then

this greatly extends how much coding you

have to do. And it becomes this kind of

really weird vicious feedback or

virtuous feedback. Not really sure which

one you want to call it where while you

code, you design and while you design,

you are coding. And so it just becomes

this kind of like really weird

undulating process in which you kind of

comes out really good products at the

end because not only did you code it,

but you also just like thought about it.

You spent so much active brain energy

coming up with ideas. But AI kind of

takes away this part, right? So if you

just vibe code it or even just light

reviewing of your code or even just

lightly reviewing what it's doing or

even heavily thinking about which tools

it's using and how they're supposed to

be using it, this time goes down

greatly, right? But that also means this

design time, you don't get this kind of

unique way in which you get to kind of

slowly come to the right conclusions

through high amounts of effort. Instead,

you have this really hyper like, oh,

make a decision now. Hey, make a

decision now. Pull it up. Use it. Okay,

I don't really like that. Make this

decision now. It's like this very kind

of myopic way to develop. But here's the

thing is every single idea that you have

can just come into the world like this.

This is the work cue I have for all the

different testing for everything that's

going up. And all of these I can click

into and see exactly what they did and

how much damage was done. And wow. Okay,

this actually honestly made it pretty

far. Good job me. Really liking what I'm

seeing right here. It has the names of

all the different, you know, trials that

it went through. Yeah, I can even see

the state of things running right now.

You can see I still have two things

running plus one of them finalizing and

kind of coming up with the strategy.

It's like I would have never created

this for my automation, but now I'm

creating this for my automation because

I have infinite ideas. But the thing is

is this right here, this is like

everything I asked for, but nothing I

actually wanted. Almost nothing on this

screen is useful, which actually feels

really frustrating because it really is

everything that I wanted. Like I can

start the next queue in any of my

different personalities. I can make it

generate a set of tasks to do for an

experiment to see if it can make itself

play the game better. I can come in here

and say, "Hey, just do a test run. Tell

me the output of your test run. Let's

graph it out. Let's see how you did. I

can do all of this. It's all available,

but it's all wrong. It's not right.

Like, I haven't had the time in here to

really come up with something good and

compelling." And that's what I think

he's talking about right here, which is

if you have something that can just spit

out anything, all of your ideas have

room to grow. And when all of your ideas

have room to grow, it's almost like you

can just turn into this like manic just

right like just like this this this

intense state where anything you think

of can be built. And the reality is most

the things you think of aren't good

ideas anyways. And so the old barrier to

entry was actually a great place for the

uh the refiner, right? The refiner came

in and said, "Okay, 99% of your gold is

just dross." It turns out that last

little bit is actual gold. Hey, since

you had to be really slow with your

decisions, you were more able to det

like to distinguish between the crap and

the good stuff. That gold, baby. So, I

just really like how he said that. Let's

just play it once more. Honestly, just

one for the boys. What that looked like

for me was just uh almost like infinite

possibility. And that part's a little

bit scary. I was ripping tokens till

midnight or two and I got I got a wife

and four kids at home. And so I'm like

putting the kids.

>> By the way, he's not even like he's not

even really selling it to you. He has a

beautiful wife and four beautiful kids,

but those kids were like he had four

kids two and under or some nonsense.

Like he the man is living life on expert

mode. So when he's just like, "Yeah, I

had four kids and a wife." It's just

like, "Bro, don't underell it. You're

staying up until midnight, 2:00 a.m.,

and then getting up at 6:00 a.m. Like,

this is going to make you feel insane.

Go to sleep, Aaron. You have four kids.

Go to sleep.

It's okay, Aaron. You good boy, Aaron.

You're you're doing a great job." And

this is what that this is what everybody

talks about when it comes to AI is that

you have so much opportunity to just

think of whatever you want and try it

out that it just like starts to eat your

brain alive. Like there were bugs back

in the day that would make me lose sleep

or there was features or challenges that

would make me lose sleep. And that's

because dude I was doing it for the love

of the game, right? I had one task I had

to really think about and I had to solve

it deeply. I had to become so dang

familiar with it. Now it's like, "Oh,

you can do anything. You can do

everything. You can do anything you've

ever wished or wanted." And now you just

can think of everything. Just think,

think, think, think. And you're just I

feel like my brain would go crazy. I get

the sirens call of wanting to stay up

late and constantly push on these things

because it is it's it's tantalizing.

Even if it's not about building

everything, sometimes even building one

thing, right? Like, oh man, hey, I'm I'm

I'm going about I'm waiting for my kids

soccer game to start. I got nothing to

do. You know, I could kick off one

little agent job right here because

there's that one little tiny UI bug.

It's super simple to fix. I could

probably get an agent just, you know,

just just little little I know. I know

it's just Saturday. It's not a big deal.

All a sudden, next thing you know, seven

days a week you're working. You never

are taking breaks and it's going to add

up and it's going to make you feel

horrible. Putting the kids down and I'm

like, "Ah, babe. I'm going to go talk to

my computer." Um, so that that has that

has definitely happened to me. Um, I

didn't burn out on that at all. I just

started to realize this is less of a

gold rush than I originally thought

because every other developer is doing

the exact same thing. And so I thought

instead of I'm going to build 40

products cuz now I have this godlike

power, I thought, whoa, so does

everybody else and nobody's buying

anything anymore cuz they're all just

building it. So much insight right

there. I I love Aaron Francis. If you

don't love Aaron Francis, what's wrong

with you? Okay, you got to you got to

love the man. But he's Dude, this is

just such another really good point

about the AI world. And I think somehow

we've all got it in our head and even

I've gotten it in my head that dude, you

got to go right now. You got to go right

now. Like things are changing super

super fast. I can see why that is so

tantalizing this like you're going to

get left behind to BS. I always think of

George Hots, right? Gios the love LLM

hate hype article. It's just like my

favorite article of all time, which is

what I don't like is two things. One,

this constant about some window

closing or the perpetual underclass or

falling hopelessly behind. This negative

valance hype not only is not true, it's

mostly designed to make you feel bad

about yourself and move to shitty San

Francisco. By the way, it's it's SF. SF

stands for San Francisco, by the way,

where everything really does suck like

how these people claim. And two, the

straw man jumped from oh hey it's a

fancy autocomplete smart compiler better

search engine to oh like it's going to

own the whole lite cone bro if you

aren't an SF and the right parties are

going to be the flash of light in the

sky one day and you're not even going to

know what happened but everything just

changed. I'll bet you everything that

this doesn't happen and ends it with my

favorite part. The people perpetuating

this are terrible people but the justice

is that this is how they feel inside all

the time themselves. It's a truly kind

of like this destructive mentality. I I

I don't know how this has arrived or

why. Is it Hustle porn that's causing

this? Like, is it Daario telling you

that all the jobs are going to be gone?

Sama letting you know just that, hey,

I'm just going to destroy the white

collar? Don't worry about it, though,

cuz you'll have tokens, bro. Hey, it's

going to be fantastic. We're going to

make everybody lose jobs, but we're

going to make infinity money. Even

though that very premise makes no sense.

How are you going to make money, bro, if

there's no jobs? What companies? who are

they going to sell to? The whole premise

is so stupid and but it's still it

infects so many people. I can't tell you

how many emails I get with people

freaking out, right? They're just like

hell. They're drowning. They're actually

being taken under wave by wave having

the riptide slowly pull them out and

that just terror and constant just

gasping for breath while being taken

under like I get it. I get that feeling.

I get that it is it is this constant

pull to be some sort of performative

nonsense. Don't go spend infinity

tokens. Instead, listen to what Aaron

has to say cuz this next part is just

like it's just my favorite.

>> So, like, let me scope it down to one or

two things and increase my vertical

ambition on those one or two things

because if I'm just ripping out to-do

lists or, you know, screen studio

ripoffs, like so is everybody else. And

so, I've had to change the shape of my

ambition. By the way, the amount of

people I do know that are actually

making a video editor is insane. At Open

Sauce, I sat down at a table with seven

creators. Three out of seven of them had

their own personal video editor they

were making with Claude Cody. It's just

like you're like, "Okay, this is crazy.

This is absolutely crazy." from being

broad to being deep because I do have a

moat in that area and now I don't feel I

don't feel like the um crazy excitement

of there's gold out west. I feel like ah

I'm I'm like curating my specific thing

that has this big moat and so I can just

go to bed because I'm not going to see

it launch on Twitter tomorrow. Even if

it does quote unquote launch on Twitter

tomorrow, the reality is when you take

your time and you keep building the same

thing until you get it into this perfect

shape, the exact thing that you've

always wanted and it has this really

good human intuitive feel, which some of

you, I hate to break it to, your

intuition is terrible, but you know,

besides for that and and if other people

feel it's deeply intuitive, they're

going to start using it. Everybody else

is going to have to chase after how you

are doing it as opposed to you just

constantly in this weird swirling chaos

of trying to figure out how other people

have done it. And I think that's what

he's attempting to say, which is just

build the thing. Okay? You having one

more day or one better night of sleep, a

probably will improve your decision-m by

a whole like a country mile. But even if

it even if it doesn't, let's just

pretend none of that matters. you're

going to just simply be at more peace

because you're not up against this

theoretical clock that's ticking. You're

just cruising. You're having a good

time. You're enjoying what you're

building and you're building what you

enjoy and you are trying to make

something that is meaningful and good.

And I know one of the critiques is going

to be like, "Well, hey, Primary, you

what are you talking about? You don't

have a job anymore. You you just make

YouTube videos. You don't know anything.

You don't even work anymore." Well, I I

do. I I've been building this game for a

long time. It's like 30,000 lines now.

35,000 lines. some enough lines of code

in there that uh every time I make a

change, you know, I have to be even more

and more thoughtful about the things I'm

doing. I have to measure performance all

the time. I have to make sure things are

really, really good. I have a backlog of

art that has to be made and I just want

to get it out. I want to make it work. I

want to make it work on a phone and then

I want to play it. Like that's the whole

point is that I want to play it a bunch

and really make something that I think

is fun and compelling. And that is

really, really hard. And for whatever

reason, I don't know why this is, I

always feel this great pressure to do it

as fast as possible. Like I just maybe

it's been the years in corporate America

that has made me feel this way, but I

just have this great pressure that I put

on myself that I must succeed at this

because I I need to succeed in this. And

so when I when I do this, like I feel a

huge amount riding on what I do. And

it's very very tempting to be like, "Oh,

just slop it up. Just make it go really

really fast." Because obviously I

slopped this up and man, it did go fast.

I mean, look at look at this. Look at

this container right here. There is a

bar on this side, a bar on this side,

and a bar [laughter] on this side. I

mean, this is just the worst interface

in the universe. But nonetheless, I feel

this pressure to succeed. And so, I've

had to also just simply learn that the

point of the task is to love the thing

you're creating. And it's something I've

just been really trying to like walk

myself through. Even though I say the

thing and I objectively know it's the

correct thing to do, it's still so easy

to bamboozle yourself. Anyways, that's

the thing I just wanted to talk about.

There's a whole bunch of other stuff

inside this video. I think you should

take the time to watch it. It talks

about like losing skills or people

losing skills or feel like they're

losing skills or people feeling like

they're having enjoyment. A lot of good

data in this video. I think at the end

of the day, the thing that you have to

focus on is what's going to be the thing

that makes you come back tomorrow.

There's this thing I I I think about

when I when I'm trying to get my swell

on, baby, which is that there is the

ideal workout, the workout you should be

doing, right? There is the one that is

going to give you the most bang for your

buck. But there's a lot of people who if

they always did the ideal, they wouldn't

do anything at all because the ideal is

just too hard. So instead, finding the

thing that makes you want to come back

again tomorrow. Some learning. If you

find that you really do like programming

and it's just something that makes you

happy, well, just make room for it first

thing in the morning, like do it for an

hour and just like love the activity,

have a little sectioned out piece of the

code where it's like, "Yeah, yeah, this

one feature, this is just my feature.

It's this the one I'm making." And then

while you're prompting and doing all the

other things, you're always like, "Okay,

okay, set up the interface for this part

of the thing to look like this. Okay,

yeah, yeah, give me the DTS file. Let me

look at Okay, yeah, make it like this

and make it like this because, you know,

when I'm doing my programming, it's

going to be exactly how I like it." And

that's great. And then you will find

that you're still liking having control

and being happy about what you're

building. Like there's there's ways

around it. This isn't some zero sum game

where you have to just be building

harnesses that build harnesses that

build programs. Okay? We don't have to

become agent herders. All right? I'm

also kind of convinced that most of

those people that are doing it,

including myself with my own testing

framework, the biggest outcome of all of

this is simply burning a lot of tokens

and giving some AI providers a lot of

money. Anyways, I just I I just I just

felt like I had to yap about it. I don't

know. I've just been thinking about

these things for a long time and I just

I just wanted to I don't know. I just

wanted to yap a whole bunch. Anyways,

really appreciate you guys. You know, I

I do miss doing a lot of uh the live

react stuff. That that was just so much

fun. I think I'm going to pick it back

up. I think I'm going to keep doing it.

Uh just because I I I I like yapping.

I'd like to keep doing it. And you know

what? My voice hurts. We're going to

figure it out. We're just going to

figure it out. We're going to do some

yapping. Okay. All right. So, I'm going

to try to make sure Tuesday, Fridays,

and maybe a little bit of Wednesday

before the standup. Awesome streams, lot

of yapping articles to cover. Okay. Hey,

be there. The name

is the primogen.