0:00
Okay. So, what we're going to do today
0:03
is we're going to be building an app
0:06
using chat
0:08
GPT and Lovable. So, if you haven't
0:11
already seen, uh, Lovable allows you to
0:15
build apps. You don't have to be a
0:17
developer. Uh, it's a no code, just a
0:20
chatbased developer, and it's pretty
0:22
cool. So, let's go ahead and get right
0:25
into it. All right. So, um, I'm going to
0:28
build
0:30
uh an
0:32
app measures
0:39
calories. Uh and the reason why we want
0:41
to do this is because if you've seen uh
0:43
this kid just made, let's
0:51
see, a
0:53
$1.12 million a month app and he's 17
0:57
years old. So, if that doesn't motivate
1:00
you or you a little
1:04
bit, then I hope this video will. So, um
1:08
that's exactly what we're going to do is
1:10
we're going to just go ahead and build
1:11
this stuff out. I'm
1:15
using as a senior
1:18
developer. Give me
1:22
the overview.
1:28
Okay, so let's go ahead and get right
1:30
into it. And this is because I'm not a
1:32
developer, right? We're just
1:35
gonna make an app that measures calories
1:38
at Cal Cali. And it's going to go right
1:40
into, of course, we're in chat, so you
1:42
might be familiar with it. I'm using the
1:44
40 model right now. And it gives me some
1:47
good, you know, instructions. It says,
1:49
"Build a custom calorie app." Okay, you
1:52
got to know your core tech. I mean,
1:54
we're going to be using AIG uh for
1:56
vision or any kind of vision for the
1:59
food recognition. Of course, we want to
2:01
have some kind of nutrition database and
2:02
then AI personalization and then user
2:04
interface things like that, right? So,
2:06
how are we going to turn this app and
2:09
make it so? So,
2:15
see, I'm going to go ahead and plug this
2:17
in right hereable. And let's go ahead
2:19
and see what it
2:22
does. Okay. And of course, here's a
2:24
bystep
2:36
road and uh I've already done a couple
2:39
of apps before but I'm just kind of
2:40
reading through this. So it says we have
2:42
to uh first define the app structure
2:45
pages to build in lovable like it
2:48
mentions the homepage the camera input
2:51
page the meal recognition result log
2:53
book the profile and this is great for a
2:55
setup uh if you haven't built an app
2:57
before and this is a little bit
2:59
confusing uh basically anytime you have
3:01
an app you're going to want to have the
3:02
different pages so that's what that's
3:03
all about there in step two um AI and
3:07
API for food recognition we want to
3:09
teach the AI or have the right
3:11
containers. So that way when the AI
3:14
recognizes the food, we're able to
3:16
properly analyze it, add it up, and
3:19
calculate the response. And uh step
3:22
three is to map the foods to nutrition.
3:24
I think that is a good uh idea. That way
3:27
we don't have to run a web search or a
3:30
web AI response every single time. We
3:32
could have a database for this. And then
3:34
uh meal logging and user dashboard. So
3:37
create a data structure uh is also a
3:40
very good step and easy to do. And then
3:43
number five would be to add some
3:46
personalization and AI feedback which is
3:48
an optional layer but I think that that
3:50
would be great. So and there's some
3:53
other things that we can do as well.
3:54
Let's go ahead and see what Lovable has
3:56
come up with so far. So you can see it
3:59
is already in the process of building uh
4:04
this app and it isn't done. It's
4:06
already, you know, built coded quite a
4:09
lot. Uh, as you can see, it's already
4:13
got to a lot of work. And this lovable
4:17
site, by the way, has already improved
4:19
so much uh since I've already started to
4:22
use it. But I can give you a little bit
4:23
better idea right
4:27
now. Okay. Okay. So, I'm going to come
4:29
back over here to chat GBT and uh let's
4:32
write out some of the exact prompts in
4:34
the steps that I need to and include any
4:36
of the vision prompts as well. So, now
4:39
Chad's going to work. Let's check back
4:41
over here with level. And that is
4:45
amazing. I mean, it it's not done, but
4:48
literally with
4:50
one prompt, and it wasn't even a a good
4:54
prompt, right? We're already able to see
4:56
a dashboard. This is highlighting. Um,
5:00
it actually works. My goodness, I didn't
5:02
think that that was all going to work.
5:04
Okay, so the the settings don't work
5:05
yet, but the home and the history do.
5:09
Uh, you know, these don't really pull up
5:10
all the way, but this is amazing. Right.
5:13
So, you can add a meal manually. Uh, we
5:16
got to get that functionality working.
5:18
You know, you can add a picture, right?
5:20
It's pulling up all my pictures. Oh,
5:23
this is incredible. Okay, so let's go
5:25
ahead and get to the next steps right of
5:29
building out cow AI. Let's see what it
5:32
uh brought me here in chat GBT which
5:36
there's two responses that we got to
5:37
choose from. So let's see
5:40
um it says lovable calorie app
5:42
stepbystep building
5:47
prompts. Okay,
5:49
so you know it already created a lot of
5:51
these
5:53
blocks. Uh,
5:56
so I'm gonna go with the one on the
5:59
right because I
6:02
think actually they're both good for
6:04
different reasons. I'm, you know, I'm
6:06
not really sure. So, I'm not really
6:09
[Music]
6:14
uh Okay, let's just go ahead and select
6:16
the first
6:18
one. We're going to just go ahead and
6:20
run this out.
6:40
Okay. Uh, and this is why some people
6:42
call it bite coding. Um, because
6:45
literally it's just keep it's just about
6:49
trying things until it works, right? And
6:52
uh that's really what coding is too. So
6:55
don't let anybody discourage you by
6:57
saying
6:58
that. But let's go ahead and get this
7:00
working. And I think it's already it was
7:03
functional because I was able to
7:08
uh press on the camera button and it was
7:11
pulling up some food. But let's go ahead
7:13
and test it. Right.
7:17
So I'm going to just pull up a meal
7:19
picture.
7:30
This is going to be a little bit more
7:32
simpler. Let's do this
7:42
one. I got the image download.
7:51
Grilled chicken salad. That's awesome.
7:54
So, uh, this one isn't a grilled chicken
7:57
salad. This is a It looks like a quinoa
7:59
or like a chicken rice bowl. So, I'm
8:03
going to press not this and see what
8:05
happens. Oh, it says no problem. Please
8:07
try again or manually add the meal. Aha.
8:10
So, let's go one more time and see what
8:12
happens if I
8:24
Can I try by manually uh doing
8:30
this? Oh, I see. So, I literally have to
8:32
put in everything, which is cool. Okay,
8:34
that's great. Just see how
8:37
far this one goes. It's going to analyze
8:40
and I
8:44
confirm and then let's see if it gets
8:46
added to the
8:48
log. Uh, it didn't really get added to
8:51
the log yet, but you can see this is
8:53
this is great, right? This is literally
8:55
tracking
8:56
um all the calories and things that you
8:59
would think it should do, right?
9:07
So, I'm going to go back over here and
9:08
all you have to do is when you have an
9:11
error is you have to just basically
9:13
press to try to fix it. It's uh not
9:17
rocket science here, folks. And the AI
9:19
is is amazing and incredible enough to
9:22
be able to build code and then also tell
9:24
you when there's an error. Okay, so we
9:26
got that error fixed. I'm going to go up
9:28
and just see if this one's still an
9:30
error. Then we can go ahead and fix this
9:31
one, too.
9:37
Okay, we got both
9:41
those says uh I need to fix the type
9:44
script error and then it fixed it which
9:46
is good.
9:51
It says there's another error here
9:56
uh or something wrong with the macro
9:57
summary and then the
9:59
index and it says that uh the AI was
10:03
able to fix it but it was not able to
10:05
apply the changes because the code
10:07
changes were the same as previous code.
10:09
So, what happens is is uh
10:14
sometimes like when I fix it, I fix it
10:16
in the wrong order. Uh or sometimes when
10:18
you press a new uh prompt, I I find that
10:23
it tends to fix some of the errors just
10:27
automatically without you having to go
10:29
and press all this try to fix it. So, I
10:31
don't know if your uh experience is
10:33
going to be the same for you, but that's
10:34
what I found out. So, let's go ahead and
10:38
take it to the next step. Okay, because
10:40
now we got the cal AI
10:44
working. And then now we can do uh food
10:47
detection. So, it it did it pretty well
10:50
already. Um, but let's see if we can
10:53
improve
11:04
it.
11:06
functionality is
11:08
working. Okay. So, what we're going to
11:10
want to do is make sure the food
11:12
detection functionality is working. Now,
11:14
it's thinking and it's going to go ahead
11:16
and do its job. So, again, while this is
11:20
thinking, I don't know about you, if
11:23
you're watching this, I'm not sure what
11:25
side of the fence you're on. If you are
11:27
a developer already, curious to hear
11:29
your comments or your thoughts on this.
11:31
Um, but if you're not a developer, if
11:33
you're brand new, if you wanted to
11:35
develop or if you thought that coding
11:37
was going to be impossible and something
11:39
that you'd never be able to do, I'm
11:41
really curious to comment down below.
11:42
What do you want to build? What are the
11:44
thoughts that are coming to your mind
11:45
right now? And, you know, what's going
11:47
to make this easier for you? And maybe,
11:50
you know, maybe I already have answer
11:51
some of these questions or maybe we can
11:53
figure that out together. And that's
11:54
really what this is about, right? All of
11:56
AI and automation is so new to most
11:58
people.
12:00
uh just by even knowing about some of
12:03
these tools, it can completely change
12:05
somebody's life. I know Chad GBT is
12:07
changing a lot of people's life, but
12:09
just like this lovable one, uh there's a
12:11
lot of other softwares that we'll go
12:13
through on this channel. So, by the way,
12:14
if you do like this content or you want
12:17
to see more stuff on AI, automation, and
12:21
how to make your life in personal life,
12:25
work life easier, better, faster, you
12:28
know, more or more results from less
12:30
work actually then go ahead and give me
12:33
a like and uh comment down
12:36
below. Okay, so right now it says let's
12:38
go ahead and implement the GPT4 vision
12:42
based on food
12:43
detection
12:45
and it was able to implement the food
12:48
detection. So it simulates it. All
12:51
right, let's go ahead and try this again
12:53
and see uh if it comes up with something
12:55
different.
12:58
Wow. So, uh, if you notice before it
13:02
just said, um, what was it? Chicken
13:05
breast, uh, grilled chicken breast. But
13:07
now it's coming in a lot more accurately
13:10
with white rice and steamed broccoli.
13:14
So, now we can confirm it's this. It's
13:16
added to the log, but it's not really
13:17
still adding it to the log, but that's
13:19
okay. Let's just go ahead and follow
13:20
along here.
13:24
Um, let's make sure that we have the
13:26
nutrition
13:33
data. Okay. And um, obviously there's a
13:35
lot of ways to do this. This isn't the
13:36
only way, but I'm just showing you how
13:38
with just two apps, Chat GBT and
13:40
Lovable, without being a senior
13:43
developer or really knowing much of
13:44
anything, we can get a working app in
13:49
very short order. And that's what's
13:51
amazing with this.
13:57
Uh I absolutely love Lovable uh and what
14:02
it's able to do because of it. I think
14:03
it's going to absolutely
14:06
change our world. And um I'm all for,
14:11
you know, giving the small guy or the
14:14
underdog the tools to beat the big guy.
14:17
Uh maybe because, you know, I been on
14:20
both sides of the fence. I think it's
14:22
great when you win. Of course, we always
14:23
like to win, but I also think that we
14:25
should let other people win when they
14:27
also have the skills. I have no problem
14:30
winning and then I have no problem
14:33
enjoying other areas of my life. I'll
14:35
put it that way because some people get
14:36
so stuck on you have to win and you have
14:38
to, you know, have this title and
14:41
championship. Um, but anyway, I don't
14:43
want to digress too much, but there
14:45
there's more important, right?
14:47
So, let's go ahead and see
14:50
as lovable gets this work done here.
14:52
Looking at your
14:53
request, you'd like to implement the
14:55
proper nutrition data and it's doing
14:58
that. So, uh it's replacing the current
15:01
mock nutrition database and implementing
15:04
the
15:05
functionality. It already integrated the
15:09
edom API which is amazing. Uh it created
15:13
a new
15:18
function. Let's see. Uh, no, I've left
15:21
the actual API called commented out and
15:23
kept the mock responses for now since
15:25
you would need to provide real API
15:27
credentials. Aha, I see. So, this is if
15:30
I want to actually take it to the next
15:31
step and make that a fully functional um
15:34
app, which I don't need to do right now,
15:36
but we would do before launching
15:38
anything like this. And then let's go
15:41
ahead
15:51
Okay. And then we're going to go ahead
15:52
and create the meal summary
16:02
page. So, as this is thinking for the
16:05
final step, um I there's we're pretty
16:07
much done, right? We can add a
16:08
dashboard, but I think you can kind of
16:11
already get the sense that this is
16:13
amazing. I mean, we can go through this
16:15
or you're more than um welcome to if you
16:19
want. I think they have a free uh
16:21
version of this. So, I'll put the link
16:23
below if you want to try levelable as
16:25
well as um I will share this actual
16:28
project itself if you for whatever
16:30
reason wanted to take uh this project
16:34
and you wanted
16:39
to There we go. Go back. uh if you
16:43
wanted to uh take this on Premiere. I
16:46
made it public so that way you can, you
16:48
know, view it and remix it. So again,
16:51
let me know down in the comments below
16:53
if you uh like this, if you have a
16:55
project in mind that you want to start,
16:56
if you think that there's something
16:58
that's a great idea, or if you just want
17:00
to share uh your feedback or any kind of
17:02
comments you have related to this video
17:04
and what kind of videos, comments
17:09
or softwares you would like me to review
17:12
in the future. Anything have to do with
17:13
AI automation and of course I have a
17:16
passion in health. So, anything on those
17:18
topics, completely open to. Uh, feel
17:22
free to leave a comment
17:23
below, like this video, subscribe to
17:26
this channel if you want to see more
17:27
videos like this, and we'll see you on
17:29
the next one.