Technology
Is Web Development Too Easy Now?
In this episode, we explore whether web development has become too easy due to advancements in AI and low-code solutions. The discussion highlights the balance between accessibility for non-technical ...
Is Web Development Too Easy Now?
Technology •
0:00 / 0:00
Interactive Transcript
spk_0
is web development too easy now. So we talked about sort of, I guess the democratization of skills
spk_0
because of AI and, you know, hey, it's cool because people who are non-technical can, like,
spk_0
be a little more technical. They can be a little more hands-on. They can vibe code things. And
spk_0
we've had a recent guest, Nathaniel, actually, if you want to go check out his episode,
spk_0
Fixing Vibed Code at Apps, I believe is the title. But he actually goes and takes things that
spk_0
someone who is non-technical, in fact, a mechanic, actually, and takes all these vibe coded apps
spk_0
that this person has made and uses and has gone through and done some security audits and other
spk_0
things like that. And you think that's cool, but now is this stuff too easy? I think this is,
spk_0
this is the theme of all these vibe coded conversations where I'm pretty sure we're actually
spk_0
going to have to hire a vibe coder and then a guy that fixes the vibe coding. So for every role,
spk_0
for every engineering role, there will be two engineers, one fixer and one person that's just
spk_0
vibing. I'm pretty sure that's where we're going right now. I mean, this is the thing though,
spk_0
right? Like, if we just think about this, like, you can't go, or in life, like, you know,
spk_0
there's good things and then you can have too much of a good thing, right? So it's like, some people
spk_0
will, I mean, not exercise, but then some people will exercise too much, right? We always
spk_0
affiliate exercises being generally a good thing for most people, depending on your health condition.
spk_0
And so it's like, you know, you can have too much of a good thing. And so like, you know,
spk_0
we like the fact that people can now play with stuff and maybe make themselves little games or
spk_0
whatever. And you know, as long as it's not harming anybody, but then there's like those,
spk_0
those like concessions or those issues where it's like people are releasing things that are
spk_0
insecure out to the public, not realizing and then people's credit cards are risk and things like
spk_0
that. Like, there's a bit of a mess. There's a bit of a mess there, right? And so like, that
spk_0
that would be an example of like going too far. And so I think you're right, Mike, in that
spk_0
there may be a vibe quarter around. However, if we just think of a positive, and I'll kick off
spk_0
the episode, I guess, with a positive, I'll hand it to you because you actually wrote these show
spk_0
notes. If we think about it, it's like, if somebody is having trouble describing an idea to you,
spk_0
and it's so non-technical that they don't even understand what you mean by, you know,
spk_0
hey, does it need a login? Doesn't need to factor. Does it need a profile screen? Is the,
spk_0
are the profiles public? And many, many other things. Those are pretty high level, non-technical,
spk_0
but yet technically related things that you need to know as a developer. Somebody might be so
spk_0
non-technical, including someone you work with, like a designer or something, or whomever is in
spk_0
in charge of like whatever goes to or what my potentially might go to market. They may be so out
spk_0
of the technical sphere that you will never be able to get that information from them verbally like
spk_0
that, but they may be able to vibe code, spend a few hours, four hours, something vibe code,
spk_0
a few different things, and then be like, I like this screen on this vibe code, or to app, I like
spk_0
that screen on this. And then for then we'll be able to answer those questions. Oh, this is a
spk_0
login screen. Oh, this is two factor. Oh, this is a third party login with Google. Oh, this is
spk_0
a calculator, like we'll be able to extrapolate that. And then we won't have that weird miscommunication
spk_0
where we make an app into like an MVP or something or maybe shy of an MVP and an alpha or beta
spk_0
or something. And then they look at it and they like, this is not what I wanted. And in fact, Mike,
spk_0
and this is where I'll kick things off for you. We had a client come to us where they wanted
spk_0
a sports website made. We were told they want a basic landing page. We made that landing page.
spk_0
I even did market research for them and things. And I thought there was absolutely nothing wrong.
spk_0
And this person came to us in like a super depressed state. It was like, this is not what I wanted
spk_0
literally at all. And this is an example of a person who is like this is very technically
spk_0
removed is not able to describe those technical basics. And so they they need something to bridge
spk_0
the gap. And so Mike, I'll let you take it from here. Oh, man, this is a mess. Like if this
spk_0
sound is interesting to you, because it probably should at this point, because it's the disaster.
spk_0
Do you want to sport the show, which you might not want to because it's the disaster. You can go
spk_0
and check us out on that Patreon. That's patreon.com slash HTML, all the things. You can also leave a
spk_0
review or rating on your podcast app, join us at your discord server or share this with your
spk_0
friends. I think I just said join us at your discord server as well. So yeah, apparently if you
spk_0
have a discord server, I don't know, I don't know what's going on here. Anyway, and if you want to
spk_0
help out yourself and us as well, actually, you can net yourself up a 20% off a scrimber pro plan.
spk_0
You can do that via the link in the show description and in the show notes, full details in the show
spk_0
notes on HTML, all the things.com. And remember that we are powered by code rabbit. They give you the
spk_0
power to AI code reviews to help cut your code review time in half. Check out the link in the show
spk_0
notes and episode description to net yourself a 14 day free trial. No credit card needed and just
spk_0
a two clicks sign up with a GitHub or GitLab account. Again, that link will be in the show description
spk_0
as well as in the show notes on HTML, all the things.com. And now I'm like that I've completed my
spk_0
part. I'm going to sign off. You do the rest of the episode. I'm going to snap. Apparently,
spk_0
that's what's going to happen today. But no, jokes aside, please, sir, take it away.
spk_0
Okay. Yeah. I feel like this might be a hot mess of an episode. So I might say a hot mess
spk_0
express. I think that's from one of my favorite podcasts out there, which is HardFork. If you
spk_0
haven't listened to them, they're sick podcasts from New York Times, definitely give them a shout.
spk_0
This might be a hot mess express. We'll see how it goes. What's this episode about? Well, honestly,
spk_0
what Matt was saying, exactly, like with all the AI co-pilots out there, the really simple
spk_0
hosting configurations, like the forcills of the world, all the advancements in web development
spk_0
frameworks that have been out there, all the build tools that had just kind of put everything
spk_0
into our lab. It sometimes feels that web development has become quote unquote easy. I don't like
spk_0
now. I just want to be very clear there. I'm not trying to get keep web development a little bit.
spk_0
But there are things about it that have become easier. And I wanted to point out that like this,
spk_0
there are these things. AI has definitely made something's easier, creating a scaffold of your
spk_0
application way, way easier nowadays, solving little problems way, way easier, right? Like if
spk_0
you're making a simple web app, that's a crud application, you know, a basic to-do list that's
spk_0
been done a million times. That stuff has been boilerplate. Like you don't even really need AI
spk_0
anymore because there's like a million different templates out there that you could just download and
spk_0
use and just adjust a little bit and have a full on crud application with the database and everything
spk_0
and login created done. I mean, 10 years ago, that was a fairly complicated thing to do.
spk_0
There was like, you know, still doable, obviously, but it would have taken a significant amount of time.
spk_0
Probably multiple developers. Now I can confidently say that I can get from zero to do list
spk_0
probably within a couple hours. That's with deployments and authentication set up all the API
spk_0
keys that I need to, you know, do without too much of an issue, right? That's where we're at today.
spk_0
But the real hard parts of web development, the actual difficult parts of web development have
spk_0
not gone away. I want to make that really, really clear. That's what this episode is about.
spk_0
It's not about the fact that it's become easy. It's because it's about pointing out what's actually
spk_0
hard and the fact that we can do the easy stuff quickly makes it so that we can actually focus on
spk_0
learning and becoming better at the hard stuff. And that's a really important skill is going deeper
spk_0
into different topics, going understanding what is actually difficult about web development is
spk_0
going to make you a better web developer because you're going to know where to focus your energy.
spk_0
Like we all know that we can pretty quickly set up a really simple application now.
spk_0
That's not the hard part. The hard parts are stuff like third party integrations, security,
spk_0
stuff like real time applications with web sockets or offline syncing, you know, doing really
spk_0
complex designs is still pretty hard and many, many more. And I'll talk about them in the next
spk_0
section one by one, but really or in the later part of the episode, but really I just wanted to
spk_0
point out that it's not about the difficulty anymore. And I think this is where advancements and
spk_0
technology always take us, right? Before, you know, the printing press and stuff like that, you had to
spk_0
go through and write each newspaper by hand, right? And that was the hard part. That was a hard part,
spk_0
not making the stories, but you had to have like hundreds of employees just writing the newspaper out.
spk_0
If you wanted to mass distribute and or like even in the earlier printing press days when the
spk_0
printing press was really slow, it's still required a ton of work. As soon as we removed that
spk_0
heart and made it super easy, what you could do is focus more employees on making better stories.
spk_0
Right? You could get people to do the hard things, the actual valuable hard things,
spk_0
rather than be buried in the minutiae.
spk_0
Well, and now AI can write. So now that hard thing has been automated, replaced,
spk_0
augmented, however you fall in that spectrum of your own opinion. But like now that,
spk_0
like to go back to the episode, now that web development had to be done by software engineers or
spk_0
web developers. And like that required a lot of self training, possibly like traditional
spk_0
education, boot camp, all these things to gain all those skills. Now we effectively have a machine
spk_0
that can help us with that or do part of that for us. Like really like what's what's
spk_0
what AI is doing is it's taking the thought process of a lot of things out. Like our machines over time
spk_0
kept getting better and better and better, obviously, right? But really there was always a human
spk_0
there sort of like doing the hard decision. Like if you put, if you took a computer, it was by
spk_0
the window and you were like, okay, like, you know, I wonder if I should wear a raincoat when
spk_0
I go outside today. And you took the webcam and you just like pointed it outside and you said
spk_0
that verbally in your room, like nothing's going to happen, right? We eventually got something that
spk_0
understands the weather, which is, you know, AI, well, I don't know if you call it AI at the time,
spk_0
I guess you did Google Assistant, which is more or less being replaced by Gemini now, but you
spk_0
would have smart speakers, Google Assistant being one of the most popular ones, right? The nest
spk_0
devices, the Google home devices, whatever you want to call them. And that was like, you could say,
spk_0
hey, do I do I need to wear? Do I need to wear a raincoat today? And it would understand your
spk_0
location. They would be just some decisions being made. It would understand that you wanted the weather,
spk_0
it would understand that you wanted it for where you were, because you know where you are,
spk_0
and those things. And that's, but that's preset up by humans. So there's some decision-making and
spk_0
now it's preset up by humans. And now we've kind of gone almost a little step further, where it's
spk_0
like these things are kind of getting a little bit more intelligent. They're able to do more than just
spk_0
say this location and this weather. They're able to make way more and more advanced decisions.
spk_0
And then, and then most recently, we've now given them agency, literally through agents.
spk_0
So like it's just these tools that are, we've taken sort of like the smarts, the smarts part, I guess,
spk_0
and we've kind of made that a machine as well, which is something that is like unheard of in all
spk_0
this, in all these years, because like I don't know about you, Mike, but do you remember when people
spk_0
were going crazy just a few short years ago, I can't even remember what a thing was called, but when
spk_0
they got Google Assistant to call that barber shop or that's the lawn to try to make an appointment.
spk_0
And that was giving, you know, a pretty rudimentary AI, giving it agency and having it say like,
spk_0
Hey, you know what? Like, I, you know, I got to leave for work at 6 a.m. I'm just going to tell
spk_0
Google Assistant that I need an appointment and the salon doesn't open till nine. So it'll know that
spk_0
and it'll automatically call, make the appointment for me, put on my calendar, maybe shoot me an email
spk_0
or text or something to let me know that it made it successfully or failed. And then that's it.
spk_0
And then like that gives it agency, right? Those things like calling the barber shop or even going
spk_0
to a website to do it, that all the intelligence of like my hair is too long, I need a haircut,
spk_0
I need to go to this hair, hair salon, I need to make an appointment, I know my own calendar,
spk_0
all that stuff was all human driven. Like our tools got better, you know, instead of writing
spk_0
things on a paper calendar, we have digital calendars and things, but the intelligence was always
spk_0
the human or almost always the human machine learning, you know, Google Assistant eventually kind of
spk_0
consumerized it, bringing into the home and now LLMs and all these other things. We're working on
spk_0
that intelligence part now. And I think that that's more clear now than ever. Because like you said,
spk_0
developers need to have a lot of skills because because making websites was hard and now it's getting
spk_0
easier, significantly easier. Yes, exactly, especially like simple websites, I should want to make
spk_0
that clear. Like for simple web apps, simple websites, it's becomes easier and easier and easier
spk_0
to make that automated to use templates, whatever. Like it's been happening even before AI,
spk_0
right? Like this wasn't just an AI thing. I think AI has definitely accelerated it to make even
spk_0
like more complex things easier. But even before that, you could say that the same things were
spk_0
happening with frameworks, frameworks making stuff easier. You say thing was happening with templates,
spk_0
like on WordPress and templating systems. So stuff was becoming easier, easier. And again,
spk_0
the consequences of that aren't that there's no work for you to do. That's never been the
spk_0
consequence of that really of these not like at least not right now in this environment. It's
spk_0
that the work changes. Right. Whether the changes, the changes are significant enough that
spk_0
you're the skill set required changes. That hasn't happened quite yet. But that's in like not
spk_0
inevitable. But that could happen at some point where like the typical programming skill set might not
spk_0
be as useful as it was before. We're not in that stage. But you know, right now what really is
spk_0
required is deeper knowledge. What's required is understanding of different systems and understanding
spk_0
the entire development lifecycle in specific cases. So I want to kind of before I get into the
spk_0
specific hard parts, what I want to do is I want to lay out what I what I mean by easy.
spk_0
What what's become easier at least from a high level to in my eyes, right.
spk_0
The number one thing is frameworks and AI generators like LLMS, right. What they've made easier is
spk_0
very simple like a framework like next JS spelled kit remix. They make interactive web
spk_0
web designs web interfaces a lot easier to create versus how they were before was was like either
spk_0
in vanilla JavaScript or you have to do some very sophisticated PHP stuff before to make it
spk_0
fairly interactive. Now with these frameworks, it's slightly easier to do those things.
spk_0
And they have a lot of patterns. That's a really important part of it too. Because those
spk_0
patterns can then be used by AI in a very efficient way to create similar things or even like you
spk_0
using those patterns to create it obfuscated to a different thing. It makes it easier for AI to
spk_0
take those frameworks and then actually make some pretty sophisticated applications with them
spk_0
on the UI side. Right. So it's just the stuff that we before again had to hand code a lot.
spk_0
For instance, if we if we talk like way before when we're talking vanilla JavaScript or even jQuery,
spk_0
just having to loop over an element and have it dynamically populate and then update. So let's say
spk_0
list of items that can dynamically change at any time. Having to keep track of what's changed
spk_0
and only change that thing or repopulate all of the items that used to be fairly complicated
spk_0
in regular JavaScript. Right. Like you had to write that entire function of like mapping
spk_0
their mapping the array to make sure that it has the same exact elements that it did before.
spk_0
And whenever it changes, you had to like react. Have custom react code reactivity code,
spk_0
not react code reactivity code that will dynamically find the element that needs to be changed and
spk_0
change only that element and leave the other ones the same. Like all of that was custom coded before.
spk_0
There were patterns you could follow and stuff like that, but that was custom code. Now it's
spk_0
literally like three lines of code in a framework. So that's the simplicity that's been coming.
spk_0
And then AI takes that to the next level by essentially automating it for you, automating writing
spk_0
that's scaffolding. That's one thing that's met that's become easy. The next is deployment and hosting.
spk_0
So before this was something that I definitely struggled with, I think Matt, you were better than
spk_0
me, honestly, at the deployment and hosting part because you had better server administration
spk_0
knowledge, but you had to have server administration knowledge to be to be able to host things
spk_0
essentially for the most part. Now in days, because of things like for cell netfly cloud flare and
spk_0
essentially one click deployment systems or deployment systems that have immediately attached to
spk_0
your GitHub or GitLab accounts that deploy like depending on where you're pushing what branch
spk_0
you're pushing to, all of that is automated for you. So you can deploy a production ready
spk_0
application that can scale to thousands of users without knowing any DevOps.
spk_0
And I'd like to say something there in particular because I mean, just balancing like literally
spk_0
load bouncing or choosing how powerful a server had to be just for an office, assuming it was like
spk_0
a larger corporation. For example, like we had to like my my old boss and I for server administration
spk_0
had to consider, hey, we might have 150 people using this. We have to see what it is they're doing
spk_0
computationally for computational load. But now that we have so many people using it, we got to
spk_0
consider the network implications of this where, you know, do we have to have a better router or
spk_0
switch, a better this, a better that just in the terms of the networking. And that's all human
spk_0
done. And like Mike was saying in that first part, the frameworks and generators, like we had
spk_0
tools, right? They were kind of like frameworks and things where, you know, you could use a load
spk_0
bouncing hardware appliance, which you know, still exists, but a load balancing hardware appliance.
spk_0
You could use different like server clusters in, you know, a hypervisor. And I'm saying a bunch
spk_0
of words that many people are listening to this. I don't know what it is. But like you could use
spk_0
server clusters and like try to load balance that way or create multiple different servers for
spk_0
certain reasons to like offload. Oh, if the person, you know, is constantly signing in, but they
spk_0
don't edit the image a lot. And the in the, you know, maybe we'll have a sign in server. So people
spk_0
can log in. I know quickly to check status of things. But then we'll have a another thing that
spk_0
offloads that heavy computational load of like re rendering a photo or something like that,
spk_0
re rendering a video. And we'll like that. If you just think about it, like what I said at the
spk_0
beginning, like that's all the like us humans using our tools are, but the intelligence is like we're
spk_0
putting these tools together. Now we're starting to get to the point where the tool is making up some
spk_0
of those decisions where it'll take a look and be like, there's these easier to use cloud things,
spk_0
like net flyversel and things like that cloud flare. So already we made it easier than what I just
spk_0
said. But then the AI can actually look at those things and go, well, I choose the choose the silver
spk_0
plan. Right. That's a totally different thing. In the past, it would be like all the things I
spk_0
just described, the hypervisors, the networking, the load balancing, all that stuff. Then it got
spk_0
easier. And it was like, okay, maybe I'll choose just for the sake of conversation, the cloud flare
spk_0
hosting plan. I'm just making this up for the sake of an audio show. Bronze silver and gold,
spk_0
which one do I need? I would still have to go and look and see, okay, it's, you know, it's able to do
spk_0
this much computation. It's able to handle this many users. It's able to store this much data.
spk_0
And then I would make the decision on bronze over gold. Now I go to an AI and I say, I want,
spk_0
I need a host and I need to do this. And I'd like it to be in this in this project range. And it,
spk_0
it does that research and reading for me and then makes an educated guess as I would and would say,
spk_0
you know, choose the silver point and cloud flare. So like, I now, I've just text prompted this thing
spk_0
and it's made the decision. I haven't made the decision. Absolutely. So yeah, there's,
spk_0
there's so many side benefits of that type of development now that it's become a lot easier to
spk_0
handle infrastructure. Let's just say that from, from the get go, like infrastructure, handling,
spk_0
whether it's due to services like cloud flare or netfly, whatever, or the fact that AI can write
spk_0
your Docker file for you and your Docker compose and all that. All of that has made it way,
spk_0
way easier to launch an application, have it scale properly for the first little while until you
spk_0
may be need to bring in an expert when you know, you have millions of users or something like that.
spk_0
I think the top level is still fairly hard. And that's going to be part of what I'm going to be
spk_0
talking about in the hard parts. But from like zero to, you know, thousands is actually become very easy
spk_0
and cost effective. Well, and that also refers to a point where we talked to our clients about
spk_0
where they'll worry, some clients are tech tech tech savvy enough to understand, like, oh my God,
spk_0
like, what if I outgrow this? And I'm like, dude, if you're out growing this, you're making some
spk_0
serious money, you're going to be in a different spot than you are talking to me today, where you're
spk_0
at minimal or zero dollars. If you're making, if this is your like little passion side project,
spk_0
you're making a fan site or something and you have a little web app or something in there with
spk_0
it to compliment it. And you're now, you know, you're now at zero today revenue and you're at 2000,
spk_0
3000, 5000 a month. And now you've outgrown your infrastructure. You're going to be able to make a
spk_0
much different decision now that you're making $5,000 more a month. You're in a much different place
spk_0
than you are today. Absolutely. Yeah. You're kicking the can down the road, but you're doing it
spk_0
on purpose. Right. I think that's the, that's the thing that you have to get through to people
spk_0
like to, to that kind of reasoning is like, yes, you might have to rewrite it. But even if we were to
spk_0
do it perfectly today, you probably will still have to rewrite something because that's just the
spk_0
nature of how software works. And many people will never even hit that as well. Which is another
spk_0
thing that makes it easier is many people might just be like, you know what? I love like the game
spk_0
Eve Online. There's huge communities of people that love Eve Online. I'm sure people listening to
spk_0
this love it. I don't know much about it myself played a couple of times. But like I know a couple of
spk_0
people that are super, super into it. They run, I think their corporations, which are basically like
spk_0
guilds and things like that. And they're super serious. They have hierarchies and they talk on forums
spk_0
and things like this. And I mean, if you're like, hey, I want to make an Eve Online site and
spk_0
there be a cool little web app where we can like manage corporations or whatever. And you know,
spk_0
you get like a niche group of people to do that. And you're making a couple thousand a month,
spk_0
a couple thousand dollars a month to someone's personal budget or a couple people's personal budget,
spk_0
depending, especially depending on where you live, is like a huge, huge amount of money. If you
spk_0
never grow it past that, you're still making a life changing impact to somebody. Like what if someone's
spk_0
just like maxed out on their budget, they can never invest. They never invest in their retirement
spk_0
or whatever. Now they make this extra 2000 and they just invest the tupel 2000. Now they can
spk_0
potentially have a retirement. Like that's literally a on a solo scale. You don't need to be a Zuckerberg
spk_0
where you hire 10,000 people and you, I don't know how many people work there, but like you know what
spk_0
I'm trying to get at is you don't need to be this big viral sensation to make a personal impact.
spk_0
If someone's handing you five thousand dollars a month in the grand scheme of like the Zuckerbergs
spk_0
and the Bill Gates of the world, that's tiny. You yourself, it's like my mortgage, my car,
spk_0
and a few other things are now free. Like that's a big, big personal impact. So kicking that kick
spk_0
in the can down the road. Yes, but also potentially removing the actual barrier period,
spk_0
depending on what you're going for. Yeah, absolutely. The last easy part that I want to talk about is
spk_0
third party services or plug-in play services, right? So if I were to do this episode five years ago,
spk_0
what would be one of the hard parts in the episode would be authentication and another hard part
spk_0
would be something like payments. Stuff like clerk and auth zero have now or even super base have
spk_0
now made authentication fairly simple to integrate. And again, something that you can rely on AI
spk_0
to help you with as well, because again, it's not AI is not creating the entire authentication
spk_0
platform is just integrating with the authentic application platform that's well known,
spk_0
which makes it a little bit safer. It's still something that I'm a little bit hesitant to be
spk_0
like full send. But if you want to implement it yourself, not that hard as well, right? Like clerk
spk_0
has very, very simple step-by-step guides, templates that you can follow to implement a full
spk_0
authentication system that you can rely on fairly heavily. It's been tested very, very heavily. They
spk_0
have a bug bounty program. You can trust clerk pretty, pretty well, same with auth zero. Stuff like
spk_0
stripe for payments, not having to rewrite a payment processor from ground up is a huge, huge
spk_0
savings in time. Now stripe isn't easy. I want to make that clear as well. Stripe has very
spk_0
significant challenges presented to it, but it's probably like 10 times easier than writing the
spk_0
entire payment platform yourself, at least. So you're still going to have edge kc problems,
spk_0
date issues, membership problems, stuff like that. But it's going to make it way easier to spin
spk_0
up an entire payment processor in your app. Another thing, databases. If you want to scale a database,
spk_0
that used to be kind of a pain. You have to know a lot about databases. Nowadays,
spk_0
super-based planet scale, neon. They all make it so that you can spin up a database in a matter of
spk_0
seconds. And again, it can auto scale for you, scale upscale down. You can pay a little bit more,
spk_0
pay a little bit less depending on usage. You can have free tiers for testing. So many ways to
spk_0
make databases easier. Same with like Prisma, which is an ORM or any other ORM that will make
spk_0
a database calls easier where you don't have to learn every single detail of SQL, like the querying
spk_0
language SQL. You can just rely on a third party that handles that in the background. You just
spk_0
need to know JavaScript essentially because it obfuscates the querying for you. So there's a lot of
spk_0
plug-and-play packages, services that have been developed and tested heavily that you can rely on
spk_0
in production today that make creating these full stack applications much, much, much easier than
spk_0
it was again, like five, six, seven years ago. But with that, we can move on to the actual hard parts.
spk_0
That still exists because again, everything that I just said, that's the quote unquote,
spk_0
easier stuff. Some of it's still a little bit difficult, but it's a lot easier than it was.
spk_0
There are hard parts that still exist that really like AI can help you with, but they can't,
spk_0
it can't fully solve. You need to be a good developer. You need to know this stuff. And if you know
spk_0
this stuff, you're going to differentiate yourself from other developers and it's going to make it
spk_0
easier for you to get jobs. It's going to make it easier for you to stand out of the job. That's
spk_0
why I want to do this episode. First point here is third party integrations. Now I mentioned
spk_0
plug-and-play packages like Cluric, Auth0, Stripe, stuff like that. That's a third party integration,
spk_0
but those are major third party integrations. Stuff that again, AI knows really well, stuff that
spk_0
has really, really good documentation, examples, but a lot of third party integrations that you're
spk_0
going to realize when you get into like the corporate world is going to be with some random CRM that
spk_0
your marketing department decided to choose that has terrible documentation that you're going to
spk_0
have to almost reverse engineer to figure out how you can make it so that your app sends the right
spk_0
data to the right place in the CRM. And then the CRM can ingest some sort of data on its own as well.
spk_0
This is where developers spend most of their time I swear is when like a random off-the-hand
spk_0
request being like, hey, I just integrated random McGee CRM with our marketing. Can we make it so
spk_0
that we can just pull a little bit of data into there in an off-handed remark in some random
spk_0
marketing meeting. And then that takes weeks to implement just those like simple things. And then
spk_0
it takes potentially years to maintain because all of a sudden you're constantly maintaining these
spk_0
really, really inconsistent APIs that don't have the proper documentation or the developers don't
spk_0
really care about backwards compatibility and they're constantly breaking like this is the,
spk_0
this is real life development maintaining these third-party integrations is a huge pain in the
spk_0
butt. And if you talk to anyone that's worked in the corporate, corporate industry, marketing,
spk_0
whatever, marketing devs, you're going to know that they're going to tell you that this is where
spk_0
all the pain points lie. Well, the thing too is like you and I have had experiences where the API will
spk_0
be let's say in the middle of an upgrade or has just recently undergone an upgrade from say
spk_0
version one to two and they'll literally say, oh, we're still working on our documentation.
spk_0
So you're kind of left trying to take what documentation there is on version two and extrapolate that
spk_0
so that it works. And then also I'd like to say this is version 2.0. So then if they go in and
spk_0
there's they start writing their document their own documentation. So you were already guessing
spk_0
how some of this works. And then they start changing things slightly for version 2.01, for example,
spk_0
because they realize that something doesn't make sense. Now we've just implemented version 2.0.
spk_0
Now there's version 2.01. Now we got to go into massage it. Hopefully there's documentation
spk_0
on this different change. So you can see how this stuff can get really muddy, especially for many
spk_0
web apps and websites that require this stuff. Like many websites and web apps, I'll give an
spk_0
example. Like if you're using something like a video game like collection app, chances are that
spk_0
person that had made that app is using an API for all the gaming information because thousands and
spk_0
thousands and thousands of video games have been released over the years. There's different
spk_0
localizations, different editions, collectors editions, special editions, all these things.
spk_0
This person did not make the app in a month and then spend 17 years
spk_0
constantly like making a database. Like they're going to someplace that has an API available.
spk_0
That's like, hey, give me all the Xbox games from 2005 to 2006 and then the API
spk_0
brings that in for them. And a lot of apps require that. Imagine a video game collecting
spk_0
app without a video game library attached to it. I mean, that's relatively useless. And so many,
spk_0
many apps get into the weeds of these APIs and these APIs are not only can they be problematic
spk_0
as we've already described, but then now they can be very vital as well.
spk_0
That's correct. Yeah. And again, like a lot of times the estimations for how to integrate them
spk_0
can be wildly inaccurate because you don't know the issues that you run into. Like the documentation
spk_0
initially could seem really simple, but then as soon as you get into like one little side feature
spk_0
that you thought was going to be super simple, but really like you have to go 10 layers deep into
spk_0
the API to pull the information and that causes a shit storm of like delays like latency.
spk_0
In your applications, now you have to think about caching and that brings up another whole
spk_0
element of complexity into your application. So a lot of these third party integrations will
spk_0
bring with it more complexity even outside of the third party integration, right, that you have
spk_0
to then accommodate and learning how to you can't learn how to you know, you're not going to be an
spk_0
expert in implementing X Y and Z integration. Well, a lot of times sometimes you can be sometimes
spk_0
if it's a big enough CRM, then you can actually like specialize in implementing that CRM into code
spk_0
and actually find work in that way. But small third party integrations that just randomly come up,
spk_0
what you have to become an expert in is reverse engineering. So just be able to read code from the API,
spk_0
from the GitHub, not trust the documentation fully, right, understand how they build the API to
spk_0
then you can utilize it to its full capacity. And you have to two, you have to be a solutions architect,
spk_0
you have to understand when you have to implement certain things, when you don't have to implement
spk_0
certain things, when you have to be able to push back on management sometimes, that's the kind of
spk_0
stuff that is intangible skills that you can develop that at least for now, AI doesn't really
spk_0
replace either the same with the other stuff that I was talking about. So it's important to understand
spk_0
that this part is is really important. The next one I have here is security. We have an entire
spk_0
episode on security. It's called we should care more about web app security. I'm going to have a
spk_0
link to it in this section, but essentially from a high level, you can think of it, you know, X
spk_0
cross site scripting attacks, cross site SQL injection attacks, a lot of different ways that
spk_0
people can attack your applications. And you can think that you're using everything correctly.
spk_0
And I've had this happen multiple times. So like some of the some of the recent reports that we've
spk_0
had come in for the applications that we're working on is one of them was the race condition attack
spk_0
where someone could submit a form twice really quickly and get two of their submissions submitted
spk_0
in something that was like a competitive competition kind of thing, right? So like that that made
spk_0
them made it easier for them to do it. They couldn't do it through the UI because we stopped the UI
spk_0
from being able to submit like it would gray out the submit button anytime they had submit. But what
spk_0
they found was like in the network requests, they had they found the API that it was sending to.
spk_0
And if they went to like postman or something, they can actually submit it as like twice or
spk_0
something like that before the rate limiting kicked in. So that was something we had to fix. And
spk_0
there's these little edge casey things like we had a similar thing happened with two factor
spk_0
authentication codes, which is a huge security issue. We had a rate limit on it, right? The two
spk_0
factor authentication rate limit. It seemed to be fine. And we had a test run against it. And it
spk_0
was fine. But what people realize is that depending on where you are in the country as like versus where
spk_0
the rate limiting was actually happening in terms of the cash that the rate limiting was using,
spk_0
you can actually use that latency to your advantage. So before the rate limit would stop you,
spk_0
you could submit like, depend again, if you're really far away from that cash, you could submit
spk_0
like 30 requests every second. And that in itself like 30 requests could potentially break a code,
spk_0
like you know, a four digit code. So we had to figure out how to fix that. So these edge casey
spk_0
security bugs that are hard to find the more you look into them, the further down the security
spk_0
rabbit hole you go, the better you're going to become at catching them before they even get to
spk_0
production. And it could have massive implications. And AI has troubles with these because a lot of
spk_0
apps don't care. A lot of open source code isn't good in terms of security. And the so the stuff
spk_0
that it's learned on is inherently bad. And therefore you're might, if you're relying solely on AI
spk_0
to secure your applications without having any eye for it, you're going to probably have a bad time
spk_0
if your app becomes big enough. Now in a small application, these things like, you know, people are
spk_0
probably not going to break it in that way because it requires some fairly sophisticated attacks.
spk_0
But the bigger your app becomes, the more you're going to need to worry about it.
spk_0
This is an example, right? Of our AI is starting to take over some human thought roles versus
spk_0
the humans having the thought roles. And what I mean by that is there are people constantly trying
spk_0
to break into stuff, find new hacks, new exploits, new ways to get into things like the 30 network
spk_0
requests, the really quick double network requests for the competitive forms. Those things, those are
spk_0
like we have humans going after that. And humans are still the thought leaders here. Right? AI is
spk_0
starting to take on a little bit of a thought leader role, maybe not even leader taking on a bit of
spk_0
a thought role. It's not quite there, right? I mean, everyone estimates it's going to get to the
spk_0
point where it's going to surpass us. And maybe that's the case and what have you, but it ain't there
spk_0
yet. And also we have crowdsourced like there's hacker groups, right? There's people that are ethical
spk_0
hackers. And God knows whether things get leaked out of the ethical hacker groups. And you know,
spk_0
who knows? But the point is is that the thought leader, the human is still on the attack for these.
spk_0
And so just relying on AI is not going to cut it. Yeah, it's not it's not enough right now. I
spk_0
don't know when it will be enough because again, the inherent problem is that it's learned on code
spk_0
that's bad. So unless someone can come in and give it enough code that's good and verified,
spk_0
what what point are you going to be able to fully trust it? I don't know. That's the biggest
spk_0
issue. Like at some point, maybe there will be a system that will audit itself, right? And it has
spk_0
enough context on how to audit itself that you could trust it. But I don't know when that's going
spk_0
to happen. That might be might be like, you know, optimistically a couple years away or maybe it's
spk_0
10 years away. Who knows? Well, if we, I mean, just like physically, like theoretically, maybe even
spk_0
philosophically to an extent, our brains are only capable of so much. If we're at some point,
spk_0
the AI who is who is helping us think will theoretically pass us because we we want to accomplish
spk_0
more than what we're physically capable of. And like, because there's obviously like people at
spk_0
varying levels of intelligence, including geniuses, you know, but but like the organ that is the brain
spk_0
has some form of limitation to it. I would imagine. So if we consider that limitation and we
spk_0
consider that AI has unlimited potential because we can always build more, build faster,
spk_0
have it build itself, had it have it learned itself, have it teach itself, have it do all these
spk_0
things. If that's if that's the case, then that's when it would be like just just theoretically,
spk_0
right? I'm not saying that there's a date or I'm even right. But theoretically, it's when the
spk_0
humans are attacking these systems, but the AI is beyond the human capacity or at the very least
spk_0
equivalent. But if you want it to be foolproof, it would be when it's above it, right? And humans
spk_0
might be able to get a quick jab in there every now and then, but now that jab will never work
spk_0
again because the AI is like, I know about that jab. Very good point. Yeah, that's exactly it.
spk_0
Moving on here for the to the next hard thing, complex CSS and design systems.
spk_0
So this one in my opinion is something that's overlooked a little bit because AI can
spk_0
Ian for you, like it can create a generic like a website for you. It can create a web application
spk_0
for you with a design. What I'm trying to refer to here more is more complex things. And second
spk_0
of all, trying to design match, let's say you have a complex design system for your corporation
spk_0
that your designers require or not even your designers, just your, you know, your the leadership
spk_0
team needs it so that all of your applications follow the same design system, follow the same patterns,
spk_0
have it so that, you know, when a user uses one application or one website from your corporation,
spk_0
it looks exact. It feels the same way as the other ones. That's a common use case. That's a
spk_0
common need in these large corporations. AI is not very good at that yet right now. Like it'll
spk_0
create something that looks similar, but to give it a one to one match, that still requires human
spk_0
intervention quite often. Complex animations that are very unique and, you know, designed by someone
spk_0
else other than an AI, also pretty difficult for AI to do one to one. So and make it look good,
spk_0
like a lot of times when I use AI to do an animation, it'll be really kind of janky in the end,
spk_0
where like it'll use a weird function that will be really fast at first and then really slow
spk_0
at the end. So you'll have to tweak that and just fine tuning is going to almost take you as much
spk_0
time as just writing it yourself a lot of times with design. Like I don't really trust it in any
spk_0
capacity at this point for design. So I think I still put design and complex CSS under the hat of
spk_0
heart things that you need to learn if you want to be a good front end developer. It's and I
spk_0
don't know when that's going to go away as well. I think the thing with some of that stuff too is
spk_0
like especially with design systems is some of it needs to be massaged. Like I like some things are
spk_0
followed in a very rigid way, but then there's going to be weird use cases where it's like hey,
spk_0
this menu isn't really defined in the design system or like hey, something, you know, has come up
spk_0
when we need to quickly spin up this web page and we can follow the design rules, but we have a
spk_0
couple of like strange like like forms or tables that maybe are like way too wide, but they they
spk_0
must be on the screen at all times. And so it's like, well, this design system doesn't, you know,
spk_0
fully explain how to do this, right? There's a reason why design systems and you know apps and
spk_0
everything get patched. It's like design systems will like evolve over time. And so I feel like
spk_0
that part is very lacking in AI. And we actually saw that with Tesla actually, which is weird,
spk_0
but I remember years and years ago when they were doing a factory tour or one of the gigafactories,
spk_0
they said, hey, we wanted this to be virtually all if not all automated, but a human is able to
spk_0
quickly jiggle the bolt in place. The robot puts the bolt in place, it jiggles out of place and
spk_0
it goes, what's happening? And it doesn't really understand what to do. And then it like messes up,
spk_0
you know, bolting the thing in and it doesn't, whereas a human just looks at it and just like
spk_0
gives it a shake without even a second thought and then just puts the bolt in and you know,
spk_0
in the correct way, it doesn't cross the writing or what have you. So I feel like that,
spk_0
this is like a virtual version of that, that little adjustment on the fly that humans understand
spk_0
and machines still do not. Very true. That's a really good analogy actually. I think it matches
spk_0
exactly to what to what like the issue with AI is right now. And which the issue with robotics
spk_0
in general is like, I think we're still far away from it being able to replace a human because
spk_0
of those like real or minutia details that is super easy for us, but yeah, kind of a rabbit hole
spk_0
to go down for a automated system. Yeah, like imagine like a bartender like has a disaster where like
spk_0
like they spill a bunch of stuff a little over the floor that also like takes out the network and
spk_0
the the POS system. It's like, okay, that person didn't plan on this when they were coming into work.
spk_0
An AI system might be like, POS is down, mess on floor, no network, can't find the broom.
spk_0
And it's over. Humans will be like, find a broom, can't find a broom, run to home hardware,
spk_0
go get a broom, we're going to get the paper out, we're going to do no like formal POS, we're going
spk_0
to do that. We'll try to do cash only or what have you or IOUs or whatever, we're going to find a
spk_0
way to like those ways, AI would just be like, POS or I don't know, you know, it doesn't get that
spk_0
quite yet. It's not it's not there yet. Whereas a human even unprepared would still try to find
spk_0
alternatives for each of those issues, clean the floor, get the payment up, etc. Absolutely.
spk_0
To save some time because there is quite a few of these, I am going to have like 10 of them in
spk_0
the show notes. I'm only going to cover probably like three or four four or five, I should say.
spk_0
So definitely check out the show notes for this episode because it's going to have even more complex
spk_0
problems that we should you should learn how to solve or you should go down at least have an idea
spk_0
on what they are. But the next one I do want to talk about and it's really important is complex
spk_0
debugging. I think AI has made debugging easier now in days because a lot of times you'll just take
spk_0
the error message, copy, paste it, put it into the, you know, whatever the agent or the chat GPT
spk_0
and ask it what, how to fix it. But the problem becomes when they become
spk_0
too complex and the error messages are too vague. And a lot of times the error messages are vague.
spk_0
Like I've had to debug error messages that literally said error and no other indication of what
spk_0
it is. And so I had to understand the system that was being developed and to end sort of to know
spk_0
where I could put in an extra console log to figure that part out, to figure out what is actually
spk_0
causing that error. And those kinds of debugging processes while getting easier with AI, I want to
spk_0
make that clear. This is something that AI is helping with. The more complex it is, especially if
spk_0
it's like a production bug that's not reproducible in local environments, those are really,
spk_0
really important to like know how your how your systems work, know which logs to look at,
spk_0
know how to quickly iterate, how to reproduce, right? Like the most important thing to do when
spk_0
your debugging is reproducing. And that requires a lot of, it's not something that you can really learn
spk_0
in my opinion, like from reading a book, it's something that you have to just be in the industry for.
spk_0
Enough, if you're in it enough, you will, you will end up looking at them and you'll end up having
spk_0
a process for how you solve them regardless of what the bug comes out. Like you won't freak out,
spk_0
you'll just follow this process of like, check these logs, put a console log here, try to serve it a
spk_0
different way locally, try to reproduce it locally this way, this way, this way, this way, this way,
spk_0
this way, this way. And that can make it a lot easier for you to get to the end goal versus an AI,
spk_0
which is just going to have to literally, in some cases rewrite your entire applications to be able
spk_0
to fix a bug that could have been, could have been fixed in like one line. Next thing here is data
spk_0
and scaling. So we mentioned, I mentioned that a scaling like past a certain point, these frameworks,
spk_0
these, you know, you know, Versel hosting environments are really good up to a certain point,
spk_0
but when you get to, again, millions of users, potentially, it's not that they can't handle it,
spk_0
they can, but the costs become so prohibitive that it makes sense, it makes more sense to start
spk_0
bringing team members on that can help you build out the infrastructure that have the expertise
spk_0
to start to, like to start to scale it properly, right? So having the knowledge of when
spk_0
to do that, how to do that to a certain degree, like you're not going to be able to know all
spk_0
of these things, but knowing that it exists and knowing when to start bringing people on to help
spk_0
you with the scaling isn't an important factor in, you know, utilizing the tools correctly
spk_0
without burying your, your project in costs and potentially latency and, and in ability to serve
spk_0
the, the actual final product. Next thing I have here is marketing and growth. Now this one isn't
spk_0
really a technical thing, but it's a very difficult one. Matt and I had a conversation today where we're
spk_0
like some of the stuff is working, some of the stuff's not working, it's very, very difficult. This
spk_0
part is, I don't know if it'll ever be fixed by AI, I know again, AI can help in this case,
spk_0
but knowing how to properly market, knowing how to, you know, build a user base, knowing how to
spk_0
make sure that your application is having the best SEO and is being served to the right people,
spk_0
networking with people like this is a very human thing in my opinion. This is something that,
spk_0
you know, you have to be on the ground, you have to be talking to people, you have to be presenting,
spk_0
you have to be doing certain, like different things, it's hard. This stuff is hard and even with
spk_0
years and years of experience, we're still screwing up on a consistent basis. So it's just
spk_0
something that depends on what you want to get out of it. That's another thing is like,
spk_0
if you're only a technical person, you might not need to worry about marketing growth, but if you're
spk_0
a entrepreneur, if you want to build your own applications, if you want to make it up the
spk_0
latter, up the corporate latter, then some of this stuff does apply to you.
spk_0
Well, the thing with marketing and growth too is it's again, one of those like human adjustable
spk_0
things where it's like you're marketing to other humans and those humans are, you know, they're,
spk_0
they're what they want to buy, what they want to look at, what they want to read,
spk_0
what they want to listen to, watch those things change over time. And also, how do you like
spk_0
convince these algorithms to show your stuff off in the case of social media and things like that?
spk_0
It's not like a simple thing. And I think like the AI's are really kind of almost taking note from
spk_0
many of these content creation platform guides. So what I mean by that is you go into something
spk_0
like a YouTube studio, YouTube has ways and like in information on how to make your site go or
spk_0
make your YouTube channels grow and how to make your videos good and you know, thumbnail game and
spk_0
all this. And that's a very textbook way of approaching it. And, and you know, they're valid,
spk_0
they're valid tips and tricks 100%. However, if everybody decided to follow all those tips and
spk_0
tricks, that textbook, if you will, of advice, not everyone's YouTube channel would succeed.
spk_0
There's still that weird little, I need to jiggle this bolt into place human element of marketing
spk_0
and growth. And, and that's an issue. Like, I mean, if you do all the things that are required of
spk_0
you, but you're kind of boring, then that's it. Right? Maybe that maybe that maybe that that's
spk_0
not good for you. But then for someone else who's very boring, it might work perfect for them
spk_0
for some reason. And that's that little human thing where we don't know what, you know,
spk_0
some of those you can't tell what exactly will work or what won't. How many times have,
spk_0
you know, we uploaded a meme or a clip or whatever and like this one's going to do great and it
spk_0
does horrible. And then with some random one that we just scheduled a week later, it was amazing.
spk_0
And I forgot about it because I didn't think it was anything special. I mean, it's happened so
spk_0
many times and you know, marketing and growth is just one of those things where the AI is going
spk_0
to regurgitate that textbook of information from things like YouTube studio. And make sure you're,
spk_0
you know, you're on camera, make sure your camera looks okay, make sure that the video has a good
spk_0
title, make sure that you have post on social media. And it's like those are valid points,
spk_0
especially for a beginner. But, you know, humans make it stuff for humans being distributed across
spk_0
machines. It's very, it's very weird. And it's very unpredictable. Absolutely. The case with that,
spk_0
I think I'm going to end the hard part section here again. Highly recommend you go to the show
spk_0
notes. So this one there's going to be plenty of information there about what the next hard parts
spk_0
are. It's at hdmlthings.com. And you can go to the podcast section there and find the show notes.
spk_0
But the last thing I do want to talk about is why the shift is a good thing. And I've mentioned it
spk_0
a couple times throughout the episode. But I just want to give the high level what I think is the
spk_0
high level important parts. We can get to the real problems. The harder things, a lot faster nowadays,
spk_0
because the easy stuff is done for us in a better way. I think that's a very, very key one I taught,
spk_0
I told, I told it said it many times. But being able to work on these hard parts that we talked about
spk_0
is kind of a privilege without having to be like again buried in a bunch of really small things like
spk_0
reactivity code and all that stuff. It's really cool that we get to fix hard things. And yes,
spk_0
it's going to become harder. As AI gets better, we're going to have to fix harder and harder problems.
spk_0
But that's awesome. You get to do hard things at work. And yes, this hard stuff might take a long time
spk_0
too. That's good. You get to get to become better at the skills that you've been learning
spk_0
without having to do repetitive tasks. The other part, a lot of this actually removes
spk_0
gatekeeping around trivial tasks. I know a lot of people that have been in development that
spk_0
are just like, I really hate this stuff because I just have to especially agency workers because I
spk_0
have to do the same thing over and over and over again. Why can't I just get this automated for
spk_0
me? You can for the most part now. If you have a good workflow, a lot of that can be automated.
spk_0
And you can just focus on the harder things. And developer developers again can focus on building
spk_0
better software like more complex things, right? Rather than building the next to do to do
spk_0
the app, which has been done a million times, maybe we can start building something much more
spk_0
complex and much more interesting. So the stuff that's going to be coming out to the consumer
spk_0
should become better. It should become more complex. It should become, it should be trying to solve
spk_0
new problems that we could not solve before because we were wasting a lot of time doing the initial
spk_0
setup, right? Like you should be focused on the harder things now. So I think that those are the
spk_0
thing, those are the reasons that I think the shift is good. I'm sure there's many more. If you have
spk_0
them, again, let me know it HTML everything on X, HTML things on pretty much all the other platforms.
spk_0
We'd really like to hear from you. But that's kind of it for me. I don't really have much else
spk_0
I really encourage you, encourage everyone to embrace that the new baseline is a little bit higher
spk_0
and to become better at the harder things. I mean, I personally, I'm going to take a bit of a
spk_0
different take on that. And that is actually that like I feel like that's a bitter pill to swallow
spk_0
for people learning right now or looking to learn. We talk so much about how things are like
spk_0
are very much well, that it's like, it's like the finish line for saying I'm going to become a
spk_0
junior developer, right? Let's just say I'm a aspiring developer and my end goal, my
spk_0
finish line is a junior developer, right? I feel like that race has gotten longer. And the more
spk_0
that we keep pushing, it's like, oh, we're only going to be working on security and things like
spk_0
that. I think it's an issue because as a someone who's like learning, you're not just going to
spk_0
learn these harder things. You have to learn those easier things. And now more of your knowledge
spk_0
is effectively like kind of whisked away. So it is a bit of a bitter bitter bitter pill to swallow.
spk_0
Is this better? Is it worse? I don't know if I'm a downer, but I kind of feel like it's not
spk_0
better. I don't know if it's worse, but I would almost argue it's maybe not better. I don't know
spk_0
what this I don't know if that's bad or whatever, but just just to me, it's like we're just pushing
spk_0
that finish line, pushing it, pushing it, pushing it. And so we tell the we tell these people,
spk_0
learn the three pillars, learn a framework, now learn the back end, now learn the servers.
spk_0
By the way, all this knowledge, AI knows that. This is all this is all garbage. Keep learning,
spk_0
though, keep learning, though, you're two years in cool. Keep learning, though, keep learning,
spk_0
though. And I just feel like now we're starting to get into like where the like the very senior
spk_0
or very talented engineers were. And we want the regular engineers to be that way. Now maybe that's
spk_0
just how humans operate. And like people are just, you know, we use our tools and we go further,
spk_0
and we go further. It's like, I don't do I didn't do the electronics engineering to build my own
spk_0
computer. I just put the parts together and understand how to configure it way, by way back in
spk_0
the day, you used to buy a home computer and many of them were you had to build on yourself.
spk_0
Right. So like, you know, the bar or the barrier to entry is different. Maybe that's this and
spk_0
it likely is. But I do think it is a bitter pill to swallow for someone who is an aspiring developer.
spk_0
It's always been like this, though. It might have accelerated recently,
spk_0
but it's always been in the same in the same breath. Like before we were, you know,
spk_0
we're always standing on the people before us. Sure. Yeah. Right. Like with with the technology
spk_0
that they build, the tool, the new tools that they build, the frameworks, the new languages,
spk_0
like, you know, when we first started, we were programming in binary and then we were, you know,
spk_0
we did assembly and then we did C and then etc. Like punch cards and all kinds of stuff. Exactly.
spk_0
Like we're always standing on the on the shoulders of giants before us where we're doing probably
spk_0
way more advanced things than they were doing back then. They had it harder than us before.
spk_0
Right. So we're always trying to we're always pushing that envelope. Again, I agree with you that
spk_0
it's a little bit of a bitter pill to swallow for people that have been learning. But if they're
spk_0
learning now, they will be learning things more advanced because they have these systems like the AI
spk_0
systems to help them learn. If we had that, we would be learning stuff way more advanced as well.
spk_0
Faster, right? When we were growing, when we were going through this. So we could get to
spk_0
the hard parts quicker and that's where they're going to be. So they're going to be potentially
spk_0
better developers than us. I hope a lot faster than we then when we got there. So that's the,
spk_0
again, I'm doing the optimist side. I think it's good to have the other side as well. But I'm
spk_0
hoping that it leads to the same ending where they're just going to be better than us.
spk_0
Well, I see the thing with the thing with that and like maybe this is its own episode. But like,
spk_0
the thing with that is like back in the day when we were learning, like doing assembly and things,
spk_0
our actual systems were very, very simple. Like punch cards and all these things. They were like,
spk_0
you know, for financial institutions trying to run huge numbers and things like that. And the
spk_0
average consumer wasn't going to websites and things. And I understand that like, you know, things
spk_0
are actually like someone going to a website today from, you know, a device in their pocket
spk_0
meaning their phone is a rather complex thing. There's a whole bunch of infrastructure that
spk_0
controls the internet. There's a whole bunch of infrastructure going into delivering the
spk_0
delivering the internet to your phone. There's a whole bunch of infrastructure,
spk_0
you know, or the whole bunch of infrastructure that created the browser. Like I understand that.
spk_0
Like you're definitely standing on the shoulders of whatever. But my, I think I think the thing that
spk_0
sucks here is that what are websites doing now more in this latest sprint? Like we're talking
spk_0
about the AI sprint. Like AI comes in and it's like, you've pushed that barrier. What are websites
spk_0
doing now that they couldn't do three years ago when AI wasn't as popular?
spk_0
Well, they're doing AI things more. And like, like analyzing based on using AI, I'll
spk_0
I'm just, but the point that I'm also trying to make is that we don't know. We're still too early
spk_0
in the in the cycle to know what they're going to be doing. That's fair. That is in 10 years,
spk_0
right? Like, well, I think the point of this is that it should push people to do the harder things
spk_0
and then create better things at the end of the day, right? Like we never, we didn't have these
spk_0
web apps that can do everything before, like, you know, 10 years ago because of the frameworks,
spk_0
these web apps became more popular. Right. And now because of AI, something else will become more
spk_0
popular. I don't, and we don't have an answer for that yet. And I feel like this is maybe why we,
spk_0
we, a lot of people see AI as a bubble because like this, this is very reminiscent at a, I mean,
spk_0
not a one to one level, but of the dot com where it was like, everyone's going to buy pet food online.
spk_0
Everyone's going to buy desks online. Everyone's going to buy, you know, like keyboards online.
spk_0
Everyone's going to, everything's got a dot com, you know, keyboard dot com and desk dot com and
spk_0
camera dot com and like watches dot com. And like, if you want to hire people named Matt, you got
spk_0
Matt dot com and like, you know, it's just ridiculous. And I feel like like AI is not just a fine,
spk_0
like if it doesn't find its place, then I wonder where we're going to be at. Because like,
spk_0
we'll say like the web apps I used five years ago are very similar to the web apps I use now,
spk_0
except now it keeps bugging me in the bottom right corner that has an AI assistant.
spk_0
And now it's even harder for me to get to support staff because an AI assistant constantly is like,
spk_0
let me help you with that and has never, ever, ever for me personally, mostly because I'm a
spk_0
researcher and I do a lot of research when I, before I reach out to support, but it has never
spk_0
assisted me. Now, maybe I could change my own workflow there and not research as much,
spk_0
and just go right to support. And then the AI would help me do the research or whatever, whatever.
spk_0
I'm not, this is why I'm saying why I'm not necessarily saying it's worse. I just don't know if
spk_0
it's better. Yeah. And I think it's a good point. I think that's it. That's it. Yeah. Yeah.
spk_0
I'm, I'm flabbergasted. I didn't think the end was going to be a debate, but that's, I mean,
spk_0
that's good. I remember if you want to have a support episodes like this, you can do so on Patreon.
spk_0
That's patreon.com slash HTML, the things and many thanks to our three dollar tier patrons,
spk_0
Tim from the web hacker on the web hacker.com Jason from geeklife radio via geeklife
spk_0
rater.com and Garrett's a goal. Level up financial planning via www.levelofanancialplane.com
spk_0
mega from guest web via yes web.sc syntaxify from the HTML, all the things discord server and
spk_0
stacy mosler from the website swoon were the designs.com. And remember that we are powered by code
spk_0
rabbit. You can give yourself the power of code AI code reviews by using code rabbit to help cut your
spk_0
code review time and have check out the link in the show notes and episode description to net yourself
spk_0
14 day free trial. No credit card needed and just a two click sign up with a GitHub or Git lab account.
spk_0
If you're looking for courses, we got that too. You can net yourself up to 20% off a scrimber pro plan.
spk_0
And you can do that via the link in the show description and in the show notes and full details on how
spk_0
to use that link are in the show notes on HTML, all the things.com and last but not late last but not
spk_0
least like the show to Michael or Raka contributing author on HTML all the things.com.
spk_0
Michael is the author of self taught the ex generation blog at self taught txg.com.
spk_0
Feel free leave a comment or a review in the platform that you are listening to this on and this
spk_0
is so true. We'll sign us off.
spk_0
You've been listening to HTML all the things podcast web development web design and small business.
spk_0
We hope you've gotten some useful and practical information from this show and we hope you
spk_0
appreciate that we talk to you like human beings and we hope you had some fun. We'll be back soon
spk_0
but in the meantime hit us up on social media on Facebook, Instagram and Patreon at HTML all the
spk_0
things and on Twitter at HTML everything. Until next time this is HTML all the things signing off.