Full Transcript

·YouTLDR

Make Unlimited YouTube Shorts With AI (FREE Workflow)

1:35:45EnglishTranscribed Jul 25, 2026
0:00

created a brand new workflow for you

0:02

that is 100% for free that you can

0:04

create unlimited viral shorts whether

0:07

it's for YouTube, Tik Tok, Instagram,

0:10

wherever you desire. And the beauty here

0:12

is you're never staring at a blank

0:13

screen. I'm going to give you a custom

0:14

GPT that you can use for free. Yes, it

0:16

works for free chat GPT accounts as well

0:18

that you can literally copy and paste

0:21

your way to creating viral videos that

0:23

is unique to you, customized for you,

0:25

personalized for you. And the beauty

0:27

here is that you're never staring at a

0:29

blank screen where even if you don't

0:31

know what to do. You can just give an

0:33

example video and then it will take the

0:36

stylistic elements, the choices of the

0:39

ambiance and the characters and all that

0:41

and creates a brand new video that is

0:44

unique and that is also having the

0:47

greatest copyrightiting hooks and

0:48

everything built in that I created for

0:50

you. All can be done 100% for free with

0:54

no monthly fees, no credits, any of none

0:57

of that hidden upsells and uh hidden

1:00

credits and throttling that happens with

1:02

other AI tools. All you need is video

1:04

express.ai and you can create unlimited

1:06

videos like this. Again, without any

1:09

hidden credits. That is super powerful.

1:11

So, let me show you exactly how to build

1:13

this entire workflow. Let me also show

1:15

you some examples and we'll go from

1:17

there.

1:19

You can see the video in your head, but

1:21

making it real that can feel like a

1:23

second job.

1:28

Adscape butt idea. It's the scripting,

1:31

shooting, re-shooting, and editing

1:33

standing between you and the finished

1:35

video.

1:38

>> Describe the story, the format, the

1:40

pacing, and how I wanted every scene to

1:42

feel. The video scene by scene, a new

1:45

character every 10 seconds with one

1:47

story flowing through every transition.

1:54

Change a line, adjust a shot, refine the

1:56

emotion. You keep creative control

1:58

without rebuilding the whole production.

2:04

And that you mount believing it trapped

2:06

in your head. Make it real with video

2:08

express.ai.

2:13

5 miles later. Lightweight, no chafing,

2:16

full range of motion. Honestly, the best

2:19

run shorts I've worn. Video Express got

2:21

it right. The best part, I forgot I was

2:24

wearing them. No stopping, no adjusting.

2:28

I could just run.

2:35

All right. So, how do we create a

2:36

structure like this? Okay. So first let

2:39

me show you this so you can kind of

2:41

understand in a visual manner what's

2:42

happening and then you'll kind of give

2:44

you a better understanding. All right.

2:45

So in this video uh if I show you let me

2:49

actually show you the actual video here

2:50

so you can see it as well. Um and then

2:53

you'll understand how we are uh doing

2:55

this properly. So there's going to be

2:57

multiple sequences in the video that are

2:59

changing along the way. All right. So

3:01

for example what's happening here is

3:04

that let me show you this example here.

3:07

um for uh instance. So when we go into

3:10

these videos, this were all created by

3:12

video express by the way. So I'll show

3:13

you what was the unique thing throughout

3:15

the process and it's each scene is

3:18

becomes its own sequence. Okay, that's

3:19

the general idea. All right. So let me

3:22

show you that right now. So let me pull

3:25

that up so you can see it. All right,

3:27

cool. So I'll mute the audio so you can

3:29

uh see it better. So all right. So okay,

3:33

so what's happening here? So you have

3:35

one person let's say speaker one is

3:37

saying something right whatever the

3:39

speech is this is pure text to video

3:41

then it passes on something to the

3:44

person here okay so this part is also

3:47

clear right so this part that part is

3:50

clear right and then the other thing

3:52

that you'll notice okay is that then she

3:55

takes it she says something okay right

3:59

see this takes it does it and Then

4:04

basically, as you can see here, she's

4:06

going to again pass it on to this other

4:09

person. Okay?

4:11

And then the person is saying something.

4:13

He's doing something on the computer or

4:14

whatever, right? And then, okay, watch

4:18

what happens here. She's again talking

4:20

and again then she passes on to the

4:24

other person. Okay? In this case, this

4:27

other person is saying something and

4:29

then okay. And then as you can see here

4:33

the transition goes in and then it goes

4:36

to the other person. Okay. So this is

4:39

basically the workflow. Okay. And this

4:43

is all done through text to video

4:46

prompting within video express. Okay. So

4:49

the idea is I created a workflow for

4:51

this where you can take a YouTube video

4:53

structure and be able to recreate some

4:55

of these elements. Okay. And I will be

4:58

editing this thing as time goes on as

4:59

well in the next coming days to improve

5:01

this further. But this is a starting

5:03

point for you to get an understanding of

5:05

how to go about doing this. This can be

5:07

from an existing video. It can be a

5:08

brand new video you create as well.

5:09

Doesn't really matter. You can do this.

5:11

Okay. So let me show you because how you

5:13

ask it questions matters more um than

5:16

what happens in the screen per se. Okay.

5:18

So let me share my screen here real

5:20

quick. So then you can um see this. Let

5:23

me share and then you can uh All right.

5:26

Let's see here. All right. So, in the

5:29

page that you come in, uh the workflow

5:31

page, okay, so pull that up here. Okay.

5:36

Is uh basically where the logic of this

5:40

whole thing appears. Okay. So, you see

5:41

this build with GPT. So, when we click

5:44

here is what I did the GPT. And of

5:46

course, I'm going to be making it better

5:46

and better with time. So, the logic

5:48

behind this is that to create a scroll

5:51

stopping video, you don't need

5:52

consistent character. You don't need for

5:55

example to have uh one person talking

5:58

for like hours at the length. You don't

6:00

need all these things. What matters is

6:02

is the video captivating or not.

6:04

Correct. In that structure that you saw,

6:07

it was four or five different people

6:10

talking in each scene, but each scene

6:13

had relevance throughout where the what

6:15

they were talking about the script made

6:17

sense. Correct? So that's exactly what

6:20

we want to achieve. The biggest mistake

6:22

people make and I've seen this so many

6:23

times is that obviously they're like I

6:26

have video express but I don't know

6:27

where to begin is the first question.

6:29

The second question is where do I go in

6:32

to actually create videos that can

6:34

actually pop

6:37

at these uh platforms where they use

6:39

seed dance 2.0 2.5 whatever which cost

6:42

$4 every 10second video and you're going

6:45

to be spending $400 and you'll still not

6:47

be satisfied because that video isn't

6:48

going to get you any results. Okay. And

6:50

with Video Express, there's no hidden

6:52

credits or monthly fees or anything. So,

6:55

literally, you're able to create what

6:57

you desire without being worried about

6:59

the costs involved. Right? Imagine you

7:01

created a video, you're not satisfied.

7:03

$50 in credits got wasted, and now you

7:05

got to restart again from scratch.

7:06

Everything that you did was unusable.

7:08

And this is the issue with every single

7:10

platform out there. That's what makes

7:12

Video Express very special. Okay? And

7:14

for whatever reason I see people

7:16

understand the value of it will get it

7:18

is that we do it at such low margin to

7:21

pass on all the savings to you. That's

7:22

almost like a gift that we give to

7:24

people. That's the greatest charity we

7:26

can ever do is not just talking about

7:28

how great we are but actually pass on

7:29

all the savings to you that does through

7:32

optimizations servers that we have um

7:35

and all the skills that we've had here.

7:37

We are not just like a wrapper like some

7:38

of these companies where all they do is

7:40

they just take a seed dance API push it

7:42

to you and then put a markup on top

7:44

right and ultimately even that is not

7:47

enough if you don't have a proper

7:48

workflow to structure it okay so that's

7:50

the general idea of it okay so this is

7:54

the u general setup so now so let me go

7:57

in here and just say build with GPT

8:02

make it give a YouTube video and it can

8:04

structure it or you could even say Okay,

8:06

I want to create a video uh that will

8:11

show the Let me zoom this in a bit.

8:14

Okay, power of video express app. Uh

8:18

generate

8:20

me six scenes of 10 seconds each. Okay,

8:26

with text to video prompts, right? This

8:29

is the general structure and um and then

8:32

u with a cohesive story, right?

8:35

So now what happens is that it's going

8:37

to go ahead and you can do that for you

8:39

right out of the gate. You see this clip

8:41

one the problem. Clip two, clip three,

8:44

clip four, clip five, and on and on.

8:46

Correct? This is one way. The other way

8:48

you can do is you can take an existing

8:49

video and you can get it to do that as

8:52

well, which I'll show you as well right

8:54

now. But the cool thing I'll show you is

8:56

you can go here, right click and copy.

8:57

I'll show you some new updates we made

8:59

as well cuz we fixed uh one of the

9:00

issues was sometimes subtitles come into

9:02

the videos in uh video express. fixed

9:04

that issue uh recently. Just I think

9:06

yesterday or so you should see it. So

9:07

all you're going to do is you're going

9:08

to go into text to video. Okay. And

9:11

you're going to come here and then

9:13

you're going to click on this. Okay?

9:14

You're going to go click on vertical

9:16

9x6. Okay? And you're going to click

9:19

here and you're going to click on

9:21

advanced right here. And then you're

9:24

just going to paste that prompt because

9:25

the advanced prompt is everything is

9:26

doing everything for you, right? You

9:28

don't have to really worry uh about

9:30

anything. It does all the work for us.

9:32

You can even click on the advanced mode

9:33

and enhance the prompt if you desire.

9:35

That's optional. I mean really up to

9:37

you. Select the create video. While this

9:39

is getting created, what I want you to

9:41

do is uh just understand right that one

9:43

of the cool things we did as well is

9:45

that uh you can go in and you can

9:47

literally do this process at a next

9:49

level. So let me show you that as well.

9:51

So while this is uh getting loaded and I

9:53

can kind of show you these uh things

9:54

behind the scenes. So effectively we are

9:57

using the text to video and then we're

10:00

copy and pasting these things to work

10:01

alongside each other. Okay, is that

10:03

cool? So what I'm going to do is as well

10:05

what a new update we're working on. I'll

10:07

show you as well kind of secret uh for

10:09

express cuz I have a special account

10:10

that I'll show you this uh with and uh

10:13

while this is uh loading I want to show

10:14

you this as well. Just give me a sec

10:17

here. Um uh so um and uh the idea here

10:22

is very simple. Okay, always remember

10:24

when we are trying to create videos with

10:26

this stuff, the structure matters more

10:28

than anything else. Okay, so the idea

10:30

will always be that when we are trying

10:32

to create these things, we want to make

10:34

sure that every scene is cohesive to

10:36

each other. Like clip one here is

10:38

cohesive to clip two is cohesive to clip

10:40

three and on and on and on, right? So

10:42

because when we paste it, it's just not

10:44

like we're pasting some random thing

10:45

that makes no sense, right? Does that

10:47

make sense? So we just go in and we

10:49

create something right out of the gate.

10:51

So, uh, that's the basic idea of what

10:53

we're trying to achieve here. So, I'm

10:55

just going to copy paste all of these

10:56

into my, so I have my video express

10:58

account here, right? So, I'm just going

10:59

to copy paste all of these into my video

11:01

express account like so. Paste and then

11:03

come back here. Copy. Right? And then

11:05

paste on and on and on, right? You get

11:07

the point, right? So, and then what

11:09

happens is even though it's text to

11:11

video only, okay, um, basically the idea

11:15

is that just it's going to do a much

11:17

more cohesive front. And as I mentioned

11:19

before, you can do it landscape, you can

11:21

do vertical, you can do whatever you

11:22

want. Okay? Uh in this particular case,

11:24

I'm doing landscape uh landscape videos.

11:26

So you justly it takes me like a couple

11:28

of seconds, right? And if you see the

11:30

other um video that I did last week, you

11:32

can even go a step further than that and

11:35

really go um uh next step, right? See

11:38

here it says dialogues are too long.

11:39

Then you just put the advanced and just

11:41

do the advanced prompt enhancement. It's

11:42

going to readjust the things for you.

11:44

Okay? So actually this one I didn't copy

11:46

properly. All right. Click, copy, and

11:48

then right click, paste. Cool. And then

11:50

let's do that here. And then here as

11:54

well. Okay. Right click, copy. Boom.

11:57

Paste. Okay. So, what's happening here

11:59

in a nutshell uh behind the scenes what

12:02

I'll show you. And you can again do the

12:03

same thing with this vertical shots here

12:05

as well. This vertical thing that you

12:06

see, same logic applies here as well.

12:08

Let's do that too by the way, so that I

12:10

show you this workflow. Then I'll show

12:11

you how you can use an existing video as

12:13

well to replicate some cool things that

12:15

could uh make it really exciting as well

12:17

for shorts mainly because shorts are

12:19

under 60 uh seconds. You can really take

12:21

it up a notch with it. So let's copy

12:23

this. I'm doing vertical video as well.

12:25

Okay, not just landscape doing vertical

12:27

as well. And this is what I did with

12:28

this uh GPT that you can get access to

12:31

in that link that I showed you earlier

12:33

which is this link right here on the

12:34

bottom vp expressflow.

12:37

Okay. So um that's the basic idea. All

12:40

right. So now let's just go in here,

12:42

copy paste and on and on and on. Right?

12:44

So we just do it and all this is

12:46

basically cohesively done on the topic

12:48

at hand that I want to achieve. All

12:50

right. So I'll show you this as well uh

12:52

in a sec. So like quick copy right quick

12:55

paste. Not really not that hard, right?

12:56

I mean it doesn't use much of a brain to

12:57

copy paste stuff. All right. So what is

12:59

happening behind the scenes that uh

13:01

while these things are getting

13:02

generated? I'll show you. Uh see all

13:04

these are being generated for me. Um so

13:07

what is happening behind the scenes in

13:09

reality is you see these things these

13:12

things are being generated for me okay

13:14

where all the sequences are being done

13:16

like so and I'll show you the uh

13:17

demonstration of this whole thing once

13:19

it's done but the idea is that even

13:21

though it's not consistent character it

13:24

still generates videos that actually get

13:27

attention for viral videos. Why it's

13:29

extremely important you understand this

13:31

because this is where people get

13:32

confused and they don't understand. So

13:35

in a viral short scenario okay this is

13:38

very important in a viral short scenario

13:40

what matters is let's say this is the

13:42

timeline like so if I uh put it up here

13:44

so let's say this is um let's say 10

13:48

seconds let's say this is uh 20 seconds

13:52

let's say this is 30 seconds and let's

13:54

say this is um let's say 40 seconds and

13:57

on and on correct this right here

14:00

assuming this this is the video length

14:02

you need to get about 83%

14:05

plus okay of people to watch the whole

14:08

thing. Does that make sense? So let's

14:11

say out of 100 people who come watch

14:13

this thing 83 people or more should

14:16

watch the whole 40 seconds. So the trick

14:18

here is that as we add more elements

14:22

into it like different people, different

14:24

environments and stuff, it makes this 10

14:26

seconds, right? Like this element that

14:28

you see here on the top like this stuff,

14:30

if I highlight it in a different color,

14:32

it's going to make this 10 seconds more

14:34

exciting and it's going to look like the

14:35

20 seconds is a different video and on

14:37

and on. So this this rate of conversion

14:40

is going to go up a lot more. Okay,

14:42

that's the trick for shorts. This is the

14:44

best trick. Okay, and you can do that

14:46

with V Express like a piece of cake. And

14:47

the quality is insanely good. So the

14:50

trick is that you can do this with

14:52

vertical, you can do it with obviously

14:53

shorts is vertical, you can do it with

14:55

um you know horizontal for like ads,

14:57

right? Um and you can do it with like

15:00

areas where like the engagement right is

15:03

very important. Okay? And this is what I

15:06

like to explain to people is that if

15:07

you're hellbent on consistent character,

15:10

that has a specific use case. Okay? Long

15:12

form videos and stuff for shorts. every

15:15

video having a jump cut that goes into a

15:17

different sequence but what is being

15:18

spoken in each of the scenes is cons

15:20

consistent and cohesive makes sense

15:23

telling a story matters a lot more okay

15:26

so while this is our finishing I'll show

15:27

you once they're done how they look like

15:29

because at the end of the day then what

15:31

we are doing effectively is that we are

15:33

effectively creating something that is

15:35

having those jump cuts like go back back

15:37

back back back again and then it'll make

15:39

it more interesting more exciting does

15:42

that make sense so this is what we are

15:44

trying to achieve. Okay. So again, this

15:46

also works with like uh videos, right?

15:48

So for example, I have like this video

15:50

right here that you see on my screen

15:51

while those other videos are generating.

15:53

I'll show you uh then I'll merge it

15:54

together and I'll export it in video

15:56

express to show you how the videos look

15:57

like. But anyway, so this was like a

15:58

weight loss transformation video, right?

16:00

So if we play it, somebody was who's

16:02

lost a lot of weight here and so on.

16:04

Amazing. Good. Awesome. So if we give

16:05

this video and then if I give uh the

16:08

same workflow like I'll go back into

16:09

this. Okay, I'll go here and I'll say

16:11

build with GPT again. I come back here.

16:14

Okay. Right. So now if I paste this

16:17

YouTube video and I say create me um

16:21

prompts to create you know let's say uh

16:26

60 seconds

16:28

um text to video prompts for video

16:31

express let's say right and

16:35

and then we say recreate.

16:39

Since we're not recreating this video

16:41

likewise, we're going to confirm once.

16:43

Yes. Okay. Because we're not copying the

16:45

video like we're not like just

16:46

downloading the video and uploading it

16:48

again, right? So basically here you see

16:51

this upload the clip. You can download

16:52

this clip or you can upload it or you

16:55

can even paste the transcript and so on

16:57

and so forth. Okay, there's two options

16:58

here. So what happens with YouTube is it

17:01

kind of does some blocking sometimes. So

17:03

what I suggest you to do in cases like

17:05

this where the YouTube video you're not

17:07

able to get um the the video itself

17:10

let's say uh because let's say it gets

17:12

blocked by Chad GPT or whatever what you

17:14

can do is basically look at this video

17:17

and you can tell it okay where you're

17:19

going to tell it like this okay so let

17:21

me show you actually one sec um uh let's

17:24

see here it'll have a description every

17:26

video has like some description elements

17:28

here let me show you this and then you

17:30

can see it okay so weight loss

17:31

transformation and motivation The video

17:33

is about that, right? So, what we're

17:35

going to do is we're going to go back.

17:37

Okay, let me show you and then we can uh

17:39

do because we are trying to create this

17:41

six clips sequence. Is that clear?

17:43

That's what we're trying to do. So,

17:45

we'll say the video is about like this

17:48

weight loss, transformation, and

17:49

motivation. Okay? And then it's going to

17:51

do the rest for us. See this? Okay,

17:55

cool. And then it's basically describing

17:57

the whole frames. So, and then I can

17:59

just tell it like get me six scenes and

18:03

the text

18:05

uh to video prompts, right? Pretty

18:07

straightforward, right? What I did here

18:09

is because some people have like free

18:10

accounts and stuff like that. I I didn't

18:12

want to make it like super intelligent.

18:14

I just want to make it simple. So then

18:16

you see this, it does this for me,

18:18

right? And again, the logic here, okay,

18:21

is that you see I'm done making excuses

18:23

and stuff. It does the work for us.

18:25

Okay. So if we go back here to V

18:27

Express, okay, and because a lot of the

18:30

times it's like about 10 seconds or so,

18:32

right, that we can generate. So what we

18:34

are trying to do is we go back here uh

18:35

while those videos were generating. Go

18:37

to create for the AI uh and then you go

18:39

to text to video advanced mode and right

18:40

click and copy paste this. I'd say put

18:42

advanced mode to automatically enhance

18:43

the prompt as well and put vertical.

18:45

Make sure that it's vertical thing is

18:47

on. Okay? Because you want to create

18:48

vertical videos for shorts, right? And

18:50

then you can come up here and create

18:51

video. While this is getting created,

18:53

let me copy the other one and on and on.

18:55

Right? Let me show you the logic behind

18:57

it. It'll make sense for you when you

18:58

see the final output because right now

19:00

it just feels like somebody's just copy

19:01

pasting something with no aim or vision.

19:03

Right? So then we come back here and

19:06

then we uh go back here like so and then

19:10

come back and then we paste it. Okay. So

19:12

on and on, right? We can do this whole

19:13

process throughout the process and then

19:16

um and then uh get the output that we

19:18

want. I'll just do it for the sake of uh

19:20

speed. I'll quickly do these things and

19:22

then I'll merge them all uh together so

19:24

you can see it what I'm doing

19:25

effectively uh for you to copy paste it

19:28

uh these things. So this vertical thing,

19:30

all this stuff is what matters here. And

19:32

then uh we'll come back here and then

19:35

we're going to come here and then right

19:37

click, copy, right click, paste. Cool.

19:41

All right. Excellent. So these things

19:42

are getting done. So while these are

19:44

getting done, I want to show you where

19:47

the logic behind media express comes

19:49

comes into the picture. Okay. Extremely

19:51

important and um that you kind of work

19:55

with this. Okay? So that we can kind of

19:57

play around with this. Very very

19:58

important. Let let me explain this.

20:00

Okay. So we go to my ad videos. See

20:01

these videos are being generated.

20:03

Correct? So what you could do is

20:05

twofold. So one is I have if you see

20:08

here carefully I have vertical 9x6 and

20:10

then landscape 16x9 on the top. Correct.

20:12

I just have to drag and drop in the

20:14

timeline to what I want to achieve. So

20:16

for example this one is landscape. So

20:18

this these variations. So I can go come

20:20

here one two uh come back here put this

20:24

in three four they all come in order

20:27

basically right so five okay and then

20:32

six okay this is like 60 seconds or so

20:36

here or take um doesn't have to be

20:38

perfect but you get the point right now

20:40

why do we need video express why can't

20:42

it do automatically because you need to

20:44

crop certain things sometimes what

20:46

happens is that we go into the timeline

20:49

here some of these ends right these

20:52

things need to be cropped they don't

20:53

look sound that good you want to make it

20:56

cropping so the transition between

20:57

sequences is really good okay we

20:59

introduced a feature called fast cut you

21:02

see this right here which kind of helps

21:04

with it where it's going to make the uh

21:06

cut sequencing much easier when you

21:08

apply the fast cut you see where it says

21:10

uh use fast cut transitions between each

21:13

uh clip so if you check this it's going

21:15

to simplify certain things that I showed

21:16

you earlier but then you know I mean

21:18

it's a It's a question of preference,

21:20

right? And then you're going to give it

21:21

what track and then what effect you want

21:23

rotation, right? It's going to give you

21:25

the previous like this, right? Rotation

21:27

non-stop like so. And this solves that

21:30

issue of that transition. Okay? And then

21:32

I click on this and then boom. So in

21:35

here, it's going to get activated this

21:37

element right here. Okay? You see this

21:38

right here? And then that fast cut will

21:40

allow the sequencing to be a lot more

21:42

easier. So let me show you this so you

21:44

understand. Okay? So, we'll let's uh let

21:46

me uh export this and I'll play this

21:48

because some of you can't hear it

21:49

otherwise. I'll say fast cut um uh video

21:54

express example, right? Which I just

21:56

showed you here. And then I'm going to

21:58

export this. So now while this is being

22:00

done, let's also uh save this project.

22:02

And I'll say um you know 16x9

22:06

fast cut video express, right? This is

22:09

the video. Let's say we do. And then

22:11

what I'm going to do is I'm also going

22:12

to do a new project. I'm going to do

22:14

vertical because you see I remember I

22:16

did the vertical uh sequences as well.

22:18

Then I'll do the weight loss one I just

22:19

did as well because you'll see how many

22:21

I could do so quickly, right? That's the

22:23

beauty. So let's go here and I'll just

22:25

go to um media library. All of them will

22:28

appear under my media, my AI videos,

22:30

correct? So there you go. And then um

22:32

I'll get to the other uh elements here.

22:35

So if you remember it ended here. So

22:37

then if I go here, this should be the

22:39

vertical videos I created. Oh, this is

22:40

not. Let me see which one was the

22:42

vertical one. Yeah, this one from here

22:43

on. Right. So, uh because this is a live

22:46

webinar, uh and I don't want sound to be

22:48

do I'm not previewing it otherwise it's

22:50

gonna echo. But let me play this. Right.

22:52

I'm going to just drag and drop these

22:54

vertical elements in here like so. Like

22:56

so. Like so. And sometimes they might

22:58

need extra editing, right? Obviously.

23:00

So, that's why I try to make sure like

23:02

uh some of these are like, you know, if

23:04

there's some any kind of uh issues that

23:06

are there in text or whatever. One of

23:08

the things we're doing with text uh

23:09

rendering issues is like we're gonna uh

23:12

update next week our first frame last

23:14

frame where if there's any text that has

23:16

to be visible the first frame last frame

23:17

will fix that. That's one of the things

23:19

we're excited about that's going to come

23:20

soon. Um anyway so this is the other one

23:24

right and I think it should all be the

23:25

verticals right? Yeah these. Okay so

23:28

let's say this is the vertical uh

23:30

elements right again you can apply the

23:32

fast cut but still don't rely on the

23:34

fast cut because AI is not perfect.

23:36

That's the whole point of a human being,

23:37

right? Otherwise, you're not going to

23:39

add any value. Everybody's going to do

23:40

the same thing. There's no creativity

23:41

here, right? We're going to make 80% AI

23:43

to do 20% we want to add a little magic

23:45

into the mix. So, again, we're going to

23:47

do the fast cuts. Okay, this one right

23:49

here. And uh just make sure that even

23:51

with fast cuts, you may have to crop

23:53

certain things. Okay, uh when you're

23:54

previewing them, keep that in mind.

23:56

Okay, even though fast cut will make the

23:58

transition better, it it still will uh

24:01

need to be set up. Okay, again, fast cut

24:03

is active here. How do you know it's

24:04

active? this little area right here,

24:06

it's going to be grayed out. Okay, then

24:08

you know it's active the fast cut

24:09

transitions. Okay, this is one of the

24:11

biggest requested things when you're

24:13

getting shorts. Hey, I want to do that

24:14

fast cuts thing, right? So, um let me uh

24:18

click on this and first let me save this

24:21

thing and then we'll export it. Uh save

24:24

this thing. So, 9x6

24:26

Video Express example, let's say. And

24:30

you'll see it won't be perfect, but

24:31

that's the reason why we have an editor.

24:32

No platform is going to ever be perfect

24:35

even if 10 years down the line because

24:36

what's going to happen is you're gonna

24:37

you're competing against everybody else

24:39

in the world you still want to make it a

24:40

little bit unique right so uh anyway so

24:43

go to my videos is that thing done let

24:45

me just take a look okay not yet not yet

24:48

we having that exported video happening

24:50

okay so yeah please be patient while we

24:52

create your videos so this 56% being

24:54

done right now one of the videos once

24:55

that is done we can export this video as

24:57

well in the cube all right so let's save

24:59

this one and then I'll go open up a new

25:01

project so let's Save this. Go new one.

25:04

And then what we can do here is

25:06

basically I'll I'll uh do this weight

25:08

loss one. Remember this one. So because

25:10

I don't want to make sure you don't get

25:11

too many things happening in your head.

25:13

Remember this one where I said weight

25:15

loss journey one. I made it create uh

25:18

you know like the scripts for that.

25:20

Remember I said give me six scenes the

25:22

text to video prompts thing and it

25:24

created this whole thing. So I copy

25:26

pasted that. Right. So let's do that

25:27

video now. I'll export that one as well.

25:29

put it together while the other one is

25:31

being finished. Okay, so you can do

25:33

multiple projects here, right? Because

25:34

of that um effectively so all the other

25:38

magic is done behind the scenes, you

25:39

know. So let's see here which one is

25:42

this?

25:44

Yeah, this one is good. And then um I

25:47

believe this one is the last one. Yeah,

25:49

so this one is where the project is

25:54

and stuff like that too. All right. So

25:58

then we're going to save these um into

26:00

the sequence. And if there's any other

26:04

um variations.

26:06

Okay, cool. Not this one. Probably this.

26:09

Yeah, this one. This one is the first

26:10

video in the sequence. Okay. So I'll put

26:13

this one here. And I'll drag these uh

26:15

out of the way. And then uh this one

26:17

will come first, let's say. And then

26:20

second

26:21

uh would be this um this one right here.

26:25

Okay. See? And then we can just drag and

26:28

drop these. This is why we need an

26:29

editor framework. That's why we're so

26:30

excited about V Express. This whole idea

26:32

of like um not having an editor to uh uh

26:36

do everything on autopilot. The problem

26:38

is it messes up the little tricks cuz

26:41

every second matters in uh when it comes

26:43

to shorts. Okay, this is the thing

26:45

people don't realize is that uh every

26:47

second matters. Okay, so this is the

26:49

second video. So I'm going to put her in

26:52

here as well. Um the sequence cuz I was

26:54

in a rush. I kind of missed this flow a

26:56

little bit. Otherwise, it'd be pretty

26:58

easy to drag and drop these. Okay. And

27:00

then like I'm doing this, this, this.

27:03

Boom, boom, boom. Okay. So, there we go.

27:05

So, this should be pretty much all of

27:07

it. Um, then we're going to do the fast

27:09

cut thing again. All right. So, the fast

27:11

cut thing. Um, we're going to put this

27:13

one as well. In this case, I'm going to

27:15

do rotation non-stop again. You can do

27:17

whatever really. You can test it some

27:19

different ones. And then now, we're

27:21

going to do this as save project. We're

27:23

going to say 9x6 vertical uh weight loss

27:28

journey. Okay. And then click save. And

27:30

then uh we can uh export this one as

27:33

well. Well uh the other one just

27:34

finished. Yeah. Cool. So let's export

27:36

this. And I'm vertical. Uh let's export

27:40

this one. And then we'll download this

27:42

other video I just saved, right? Fast

27:44

video express example. Uh if this thing

27:47

works, that is. Let's see. Okay, let me

27:50

just download this thing cuz we're

27:52

making a lot of updates behind the

27:53

scenes for Video Express to release our

27:54

next version. So, I'm using that updated

27:57

behind the scenes version. So, if you

27:58

see some things a little bit here and

28:00

there, you know that. Okay, so Video

28:01

Express

28:03

example. Okay, cool. All right, let's

28:05

save this. So, let me open this up cuz

28:07

then you can hear it clearly for audio

28:09

purposes. So, let's play this thing.

28:13

All right, cool. Let's play this. Let me

28:15

load this up so you can see it. And it

28:17

won't be perfect. Again, like I said,

28:19

that's why we want to edit it. But I

28:21

want to show you how quickly I can cut

28:22

these things up with the text to video

28:23

feature for shorts where the transitions

28:25

are clear.

28:33

Creating videos shouldn't take my entire

28:35

weekend

28:40

idea into videos prompt.

28:43

What if one sentence became a complete

28:45

video?

28:54

It's building everything for me.

29:07

It's building everything for me.

29:16

One idea, unlimited possibilities.

29:28

Now I spend time creating, not editing.

29:41

All right. So now you saw that right?

29:43

Some elements like let me actually show

29:44

you those exact elements here so you can

29:46

uh see it more clearly. So I'll do it

29:49

this way maybe hopefully like this.

29:51

Yeah. So let's watch this carefully.

29:53

Okay. So watch this where I could have

29:54

improved upon. I'll show you on how to

29:56

do it as well in real time. See this one

29:58

was good where this guy speaks and

30:00

everything is good. He speaks it and

30:02

everything. Right. Okay. And then you

30:05

see that transition here. I didn't like

30:07

it. Right. This was too much blank from

30:09

8 seconds in. You see this? I would have

30:12

cropped this thing. Okay, so in video

30:14

express, I'm going to show you how to

30:15

crop this stuff so that it's like back

30:17

back. It should be pretty good. This is

30:19

the reason why I say AI never never good

30:22

for this. You want an editor. Okay, if

30:24

you're going to miss out on this, this

30:25

is where like people make a mess out of.

30:27

Okay, now let's play this one more time.

30:28

Again, let's go forward here. There was

30:30

another area I didn't like. So this one,

30:32

there was some speech here talking here.

30:34

Exactly here. You see this? Too much too

30:37

much transition. Okay, I wouldn't do

30:39

this this way. I would crop this as well

30:41

and go straight to this element where

30:42

she starts talking, right? As you can

30:44

see here. And again, you'll notice one

30:47

other area here where see there's too

30:49

much pause happening and then this

30:51

animation occurs. So I'll probably cut

30:53

that pause. And then right here where

30:55

she starts walking, I'd probably come

30:57

start again, right? And then you'll edit

30:59

this thing and you'll basically make it

31:00

a much more exciting video right here.

31:02

Again, as soon as this animation ends, I

31:04

would move to the other sequence too

31:06

much. The animation is happening for too

31:08

long in my opinion. Um, and then here as

31:10

well, I move it to this area where

31:12

there's talking. There's a pause

31:13

happening here, which I didn't like for

31:15

too long. You see what I'm saying?

31:16

Little little details. You want to chop

31:17

chop chop. Make it really exciting. All

31:19

right. And then there was this black

31:21

area. You watch this this area from

31:23

around 54 seconds in of this video right

31:26

here at this spot. You'll notice there's

31:28

like a black screen. What is it? And

31:30

then it does this circle thing, which

31:32

was that transition we put. It's okay,

31:34

but you know, not matching here, right?

31:36

Cuz there's no speech happening. and

31:38

then there's a fade out. So, so many

31:40

areas where you see some little bit of

31:42

manual editing helps. Correct? So, uh

31:45

this is the basic general reason why I

31:47

say we want to make sure that the

31:49

editing framework is always intact.

31:52

Okay? You want to crop these. So, what

31:54

I'm going to do um let me actually uh

31:57

explain this to you in more detail. Um

32:00

so then what I would do so let's say we

32:01

save this uh project, right? So let's

32:03

open this project that I'm talking

32:04

about. So this fast cut um 16x 91. So

32:08

this was the project right I just

32:10

explained to you just now. So what I

32:11

would do is let's say some some of these

32:13

areas where there's too much stuff where

32:15

like this blank screen area remember I

32:17

showed you. I'll just create a crop

32:19

element. So what I would do um is you

32:22

see this um element right here which

32:24

says this uh cut thing. See this right

32:27

here? So I would click on that and then

32:29

I'll chop that right away. Right? So

32:31

I'll click on this area. I'll select it

32:32

with this little purple thing. the

32:34

slider and then uh I'll just crop it.

32:37

Okay. Boom. And then you see there's a

32:39

little piece here. There's this little

32:41

piece here that's out there. I delete

32:42

that. Okay. So, I'll just uh click on

32:45

this and then I'll just right right

32:48

click, right click on it, and then

32:50

delete. Boom. Okay. Then I'll snap this

32:52

back into the frame. Okay. So, then the

32:54

transition is a lot more fast-paced. So,

32:56

if I play it, there we go. It goes

32:58

straight to the other one. Right. And

33:00

then remember I showed you this area

33:01

where like in this section I know it's a

33:04

little bit off. So I'll zoom this in a

33:05

little bit so you can see it better. Um

33:07

hopefully um yeah here. Okay. So you see

33:11

this area again from at this point in

33:13

time to this area there's a lot of dead

33:16

space. Okay. And uh again AI does not

33:19

have a proper understanding of vision

33:21

and that's where you come into the

33:22

picture. So I'll then again click it and

33:24

then I'll click on this uh crop element

33:26

here and then I'll crop it. Okay. Boom.

33:29

and then right click and delete this

33:30

piece that you saw there. Then I'd snap

33:32

this back in this next scene. Okay. And

33:34

then you see again even here when it's

33:36

starting there's so much dead space for

33:38

the first couple of seconds until uh

33:40

this the talking happens. Correct. So

33:42

then I'll again snap this back in

33:44

wherever I feel comfortable in the

33:46

speech area like here maybe and then

33:49

I'll cut and then this one that I

33:51

selected right this one right here the

33:53

speech I'll just delete it. So this is

33:54

what we mean by making it exciting.

33:56

Okay, this is the difference between

33:58

successful people who create videos

34:00

versus people who just want AI to do

34:02

everything and then they keep failing

34:03

again and again because every second

34:07

matters. Okay, you won't want any fluff

34:09

in there. We can sit and discuss how

34:11

automation will play a part but also

34:12

understand that's your edge right if uh

34:15

here. So again remember I told you

34:16

there's too much delay happening in the

34:18

sequence here. So, um, like what I would

34:20

do is there, right here, somewhere

34:22

around here, I'll make a cut because I'm

34:24

not too satisfied with the slow

34:25

transition happening. I'll cut and then

34:28

I'll go here and then, um, at this point

34:30

in time where this animation starts,

34:31

I'll cut this. Boom. And then right

34:34

click, delete. Snap this back in again.

34:36

Right? And, uh, you see now it became a

34:39

much shorter to the point video. Right?

34:42

So if you think about it, so now if I

34:43

come back here, I come into this

34:45

sequence and then let's snap these into

34:48

in here, right? And then this through.

34:50

So it's just like going to take you that

34:51

extra couple of minutes or whatever, but

34:53

like you're making it more exciting

34:54

here, right? It just go straight to the

34:56

point where um the the transition is

34:59

making sense, correct? So then again, I

35:01

can crop this stuff and then cut this

35:03

like so. And then it crops it. And if

35:05

there's any of these elements that you

35:07

don't like, you know, you can just uh

35:09

cut this out, right? just makes it a

35:10

little bit more exciting instead of like

35:12

boring stuff that you see where you feel

35:14

like there's extra pauses happening and

35:16

all that. Right? And there we go. Okay.

35:18

So now if I don't like any of this

35:20

animation where I said this animation

35:22

was too slow, right? I felt like it

35:24

should have been done right about here.

35:25

Okay. So I can just select this red one,

35:27

select this whole thing. You see what I

35:29

was doing here? I selected this area

35:31

right here. And then I just delete this

35:33

area. Okay. And boom, cut. So now it

35:36

goes straight to this and it snaps back

35:38

into it where this guy uh is showing the

35:41

stuff. He's speaking and he's talking

35:43

and then I felt like there's these

35:45

elements here where it's just so much

35:46

dead space here. It's good. I mean in an

35:49

animation perspective, uh it's pretty

35:51

good, but then I felt like there's too

35:52

much delay where nothing seems to happen

35:54

for a period of time, right? All this

35:57

stuff. Um and then he starts talking

35:59

towards the very end um again, right? So

36:02

somewhere around here, I'll just crop it

36:03

out and I'll just uh get rid of it.

36:06

There's this black screen happening. See

36:08

this? This is the other one I mentioned

36:09

to you earlier. This black screen, I

36:10

don't like it. Um it looks silly. Okay.

36:13

Again, cut. So then we come back here

36:17

and then you see that transition. I said

36:19

this transition was weird where it

36:21

started like this. It rotates that

36:22

person's face and the uh head and all

36:25

that. It comes here. All this dead

36:27

space, right? And then we come here. So,

36:30

I would probably start somewhere down

36:31

here for this sequence. And um good

36:34

thing with RePress, you don't get build

36:36

for these, right? Imagine like you

36:38

didn't like any clip and you want to

36:39

regenerate it with these other

36:40

platforms. My goodness, like going to

36:42

cost you like tons of money. All right,

36:44

cool. So, now you see this video became

36:46

from a minute long to about 41 uh

36:49

seconds. Okay, so let's just export this

36:51

one as well. Um

36:54

um let's say exciting version or

36:56

something. Okay. And then let's export

36:59

this stuff. So, while this is exporting,

37:01

I'll I'll I'll load up that other um

37:04

video that we did, that weight loss one,

37:06

right? So, let's we go here. You see

37:08

this vertical weight loss journey one.

37:10

So, let's just um load this up. And this

37:13

is a vertical video. So, playback might

37:15

be a little bit off, but it's okay. All

37:17

right. So, uh let me load this up right

37:20

now. Um okay. So, let's just load this

37:24

video in my computer and I'll show you

37:27

so you can hear it properly as well.

37:28

Okay,

37:36

I'm done making excuses.

37:50

Small choices create big results.

37:59

Every journey starts with one step.

38:09

Discipline beats motivation every single

38:12

day.

38:12

>> Let me show you something even more

38:13

exciting. All right. So, if you go into

38:16

our um into our Video Express

38:18

application, so let me show you here.

38:21

Um, and you'll get an understanding of

38:23

this even better. So, like here, we go

38:26

in here. Remember that other video that

38:28

I uh did? I'll put my screen up here.

38:31

Remember that other video? We didn't

38:32

render that one yet. We didn't generate

38:33

it obviously yet. So, let's go to that

38:35

video right now. And then that project.

38:37

And let's open that project up. Remember

38:39

this one right here, which was uh the

38:41

9x6 Video Express example. Um, let just

38:45

save this one. And this opened up,

38:47

right? This one was the vertical video.

38:48

We did not generate this. uh we not

38:50

render this out. So what I'm going to do

38:52

in here, see, same thing where there's

38:54

like sequences where like it does not

38:57

like has these uh transitions. So this

38:59

red one, we'll just crop it, right? I'll

39:01

just make it a little bit exciting so

39:02

you understand how the excited version

39:04

will appear. Okay. So this one right

39:06

here as well, the transition is a little

39:08

bit off. Um I feel like it could be

39:10

cropped to a little bit a couple of

39:12

seconds here, a couple of seconds there.

39:13

Right? And then um this one right here

39:16

as well where you see that transition is

39:19

happening in that laptop. Like there's

39:20

so much dead space here. I'll crop this

39:22

one as well. And then we come back here.

39:28

Leave with the phone text. So I'll move

39:29

to her right here where she's talking um

39:32

for example. Okay. And I'll crop this in

39:35

here. And then uh this transition to

39:38

doesn't make any sense here. Again, too

39:40

much nonsensical stuff happening. So if

39:42

you're not happy with anything, you just

39:44

cut it. All right? You understand what

39:45

I'm saying? So like uh if you don't like

39:47

anything, uh just cut it. And not

39:50

everything is going to be perfect

39:51

obviously, right? From uh what we see.

39:53

All right. So then we come here and then

39:55

you see this guy right here is showing

39:57

something on the screen. And so I can

39:59

start around this point. Okay. So this

40:02

point right here. So I'll just crop this

40:03

part right here. And you can crop it in

40:05

multiple ways. Right. So then gen

40:07

generally that beginning and the end

40:09

part is where the pauses happen which

40:11

messes up the short flow. Okay. And uh

40:14

often times right once it's bad it's

40:16

kind of really kills the conversions of

40:19

that video like the person wanting to

40:21

watch that video fully right see some of

40:23

these sequences. So I'll come from here

40:25

I'll show you how the end result will be

40:27

okay just so that you understand what

40:28

I'm talking about. So, um, and then you

40:32

see some some cases like here we're

40:33

going to, so this one right here where

40:35

sometimes it generates this stuff right

40:37

here, we're pushing an update that fixes

40:39

this where it never shows the stupid uh

40:41

text, okay, which kind of happens once

40:44

in a while. So, we're going to be

40:45

pushing that in uh probably in a couple

40:47

of days. We're doing some testing, so

40:48

that will be completely gone for you. It

40:50

won't show up. So, I'll not even use

40:52

this element at all uh in my video. Um,

40:55

so then you go into this one where some

40:59

it comes into this transition. You see

41:00

what I'm doing right now? I'm trying to

41:01

make it exciting. Basically, I don't

41:03

care about the length. I want 83% of

41:04

people to complete the video, right? So

41:07

that they want to watch it, it completes

41:09

it and then I finish the process of what

41:11

I'm trying to achieve. You can always

41:13

create more length like you can extend

41:15

that uh frame and you can always do it.

41:18

But the logic stays, right? You get the

41:20

point of what I'm trying to say. So

41:22

that's the idea of these projects when

41:23

we're building these in uh on um from

41:25

scratch and stuff. You want to make sure

41:27

that they cohesively make sense. And

41:29

then now in this case if it's 24

41:30

seconds, it doesn't really matter

41:31

because it's 24 good seconds. Okay, that

41:34

that's what matters. So then example um

41:37

fast version let's say or whatever we

41:39

can do that now and generates. Okay,

41:41

does that make sense? At any point the

41:44

GPT that I gave you, you can go back

41:46

here and let's say you didn't like any

41:48

um sequence, right? Then you can say

41:50

something like this. Okay, you can say

41:53

make sure each sequence is uh each scene

41:57

right is no longer than uh 5 seconds

42:02

long. Regenerate and make sure it's

42:06

exciting where the characters speak

42:11

right away without any pauses. All

42:14

right, it won't be perfect. That's what

42:17

reason why I still want you to edit but

42:18

you can still guide it a little bit on

42:20

this front where instead of 10 seconds

42:21

you can make it five and this is where

42:23

many times right when I try to explain

42:25

this stuff to people it kind of uh

42:28

people kind of don't understand between

42:30

the difference between length of a clip

42:33

versus length of a video okay that means

42:36

a clip's length could be 3 seconds 5

42:39

seconds 8 seconds 10 irrelevant the

42:42

video can be a minute 2 minutes 3

42:44

minutes whatever the idea is that the

42:46

jump cuts and the transitions that we're

42:48

making was what will make it exciting.

42:49

Okay, you'll see it even in Hollywood

42:51

movies. You're never going to have a

42:52

scene fixated into one uh person

42:55

forever. It's going to be too boring.

42:57

Even lengthy dialogues, okay, if you go

42:59

back um you know in time for the most

43:01

classic movies, they're going to have

43:03

different angles and they're going to

43:04

move from the reaction of a person who's

43:06

uh listening to that dialogue, right?

43:08

How do they react? All these little

43:10

nuances are there to keep it exciting.

43:12

So when people uh ask us, right? And

43:15

this is the biggest thing I want to

43:16

hopefully uh drive upon today is that my

43:21

idea of a video is 83% of the people

43:25

plus are they finishing and for that I

43:28

may need multiple scenes that have to be

43:30

merged together. Okay, it's very

43:32

important like two three because if it

43:35

was one big video of that one person

43:38

let's say this is the character talking

43:40

it loses the interest when it's a short

43:42

video. Okay, loses the interest. If it's

43:45

a short video, it's long form, no

43:47

problem. But long form audiences are

43:49

different from short form audiences.

43:50

Okay, so we try to recreate the

43:52

sequences based on this. All right, and

43:55

uh now you see this, it uh generated me

43:57

this again. Um and on and on, right? You

44:00

get the point. So you can prompt it to

44:02

get it and you can copy paste it. Now

44:03

it's about 5 seconds long. One trick you

44:06

can play is that the dialogues, right,

44:08

will make sure that it is going to be

44:10

five seconds or length. That's the idea.

44:12

So that if somebody's speaking a big

44:14

monologue and it's going to stretch

44:15

beyond 10, it's going to be a problem.

44:17

That's that's the reason why I did the

44:18

GPT. But make sure what you type into

44:21

GPT matters more than anything else.

44:22

Okay. I made sure that the frameworks

44:25

are set that it does it in a proper way,

44:28

keeping the uh excitement intact

44:30

throughout. Okay? And I'll be improving

44:31

this based upon what people are doing

44:33

with this uh workflow in this workflow

44:35

page. So watch out for that so you can

44:37

uh build on top of it and stuff with

44:39

time. What's happening is as we move

44:41

work more more towards the workflow side

44:43

of stuff where uh these chpts your

44:46

clouds and entropy cloud and you know

44:49

other platforms out there as they move

44:50

more into the uh it's doing tasks for

44:52

you stuff understand that what's going

44:55

to separate the you know people who use

44:58

AI completely to people who will succeed

45:00

kind of thing will be these little

45:02

nuanced edits right refinement we call

45:04

it okay and as these AI tools like

45:08

express get better and better and um

45:10

with time then it'll merge that gap as

45:13

much as possible where that little bit

45:14

of creativity that 2 3% of creativity is

45:17

what's going to uh make it very very

45:19

powerful okay for people okay now again

45:22

I would have copy pasted this stuff

45:24

throughout the process and um go from

45:27

there correct now let me explain to you

45:30

how the structure of this whole thing

45:32

works okay because if you don't

45:34

understand that it's almost like you

45:36

know it goes to nothing okay So let me

45:39

actually save this thing right now and

45:41

then I'll open up this sequence here so

45:43

you can uh see it uh more clearly. Then

45:46

I'll write this down as well so you can

45:48

uh play around with what I'm showing

45:49

you. Okay. All right. So the idea is the

45:51

logic behind this is that when we use uh

45:55

text to video which is under under uh

45:59

create from prompt. Okay. Text to video.

46:02

What we are trying to achieve here is

46:04

that if we say the timeline in the

46:06

bottom right see this is the timeline

46:08

right if I highlight it in a way so the

46:11

timeline in the bottom what we are

46:13

trying to achieve is it should be let's

46:15

say this is the flag that we're going to

46:17

say as the finish line right like in the

46:18

Mario level let's say super Mario level

46:21

this is the finish line right this is

46:22

the finish flag and then um if that is

46:25

the timeline what we're trying to

46:26

achieve in a nutshell is very simple so

46:29

we put like uh the clip which is one and

46:32

and then two right and then three it

46:35

doesn't matter the clip length or each

46:38

of them four how many ever is required

46:40

right and even if you do five okay

46:42

whatever the small one there the idea is

46:44

that when a person is watching right the

46:48

transitions okay here to here to here to

46:52

here what we're trying to do is what's

46:55

called a jump cut okay we change the

46:58

sequence completely we change it where

47:00

the person moves into a different

47:02

environment and that sort of thing,

47:03

right? And it it can change the

47:05

characters and stuff like that. What

47:07

happens there is for the human mind, it

47:09

feels like they're watching a completely

47:11

new video because remember for shorts

47:13

they're scrolling, right? A person in a

47:15

short scenario, if you look at the logic

47:17

behind it, it'll make a lot more sense.

47:19

Um let's say this is your mobile phone,

47:21

right? Uh of the users watching it. So

47:24

they are watching these videos like so,

47:26

right? And then they're just flipping

47:28

their phone up up up up. So what we're

47:31

trying to achieve here basically from

47:33

this logic here is that when they flip

47:35

it up right for them when they see this

47:38

right two three the sequences are

47:41

changing in their mind it looks like a

47:43

completely different video they'll stop

47:44

because otherwise it's just going to

47:45

keep flipping correct so that's the

47:47

objective and that's why text to video

47:49

works so well within video express for

47:51

this use case of shorts in particular

47:54

okay and when you apply it think about

47:57

it from the logic of what people were

47:59

watching our attention span have gone

48:01

down. Right now, the average attention

48:03

span of a normal person uh is 8 seconds.

48:06

Okay. Um now, obviously there's like no

48:09

deep research done on it, but it kind of

48:10

makes sense if you think about it where

48:12

people don't even pay attention. Like my

48:14

videos, as an example, even the best of

48:17

my videos that I've created, I lose

48:19

about 30 to 40% of the people within the

48:21

first 30 seconds. You can be the best

48:24

video, it doesn't matter. People's

48:25

attention spans are so bad for 100

48:27

different reasons, right? So when it's 8

48:28

seconds long, understand me, we are

48:30

trying to create a new gap in the brain

48:33

of the person where when the transition

48:36

happens, right, this transition here

48:38

happens from here to here to here to

48:40

here. We're trying to trick their brain

48:41

into thinking that hey, it's a brand new

48:43

video, so I should not flip it. So then

48:47

then what happens? What happens? So if

48:50

that uh variation like I told you is 83%

48:54

or more in a shorts video, very

48:56

important, keep that in mind. retention

48:57

rate we like to call it 83%

49:00

plus meaning 83% of the people finish

49:02

the whole video then YouTube Tik Tok

49:06

Instagram whatever uh whatever platform

49:08

you want to use what they're going to do

49:10

is they're going to feed this into the

49:14

when you go into your homepage of your

49:15

YouTube app uh when you're on your TV

49:17

and you're watching they will show in

49:19

the recommendations okay right now if

49:21

you come into the um I'll show you

49:23

hopefully you can see it um let Let me

49:26

see. I'll try in this uh browser so that

49:29

it'll be easier for you to understand I

49:31

guess. So yeah, let's see. If you go

49:34

here, so if you go to your YouTube thing

49:36

and I didn't log anything here, right?

49:38

So So you see this how the system works.

49:41

So you have and you'll understand that

49:44

YouTube made this change recently. So

49:45

you're going to have this right here.

49:48

Okay. Okay. So you're going to have one

49:50

video here. Second video. I don't know

49:52

why the third one doesn't show but let's

49:54

say assume third one will show sometimes

49:56

right so three three on top and then if

49:59

you come down here

50:01

look at this

50:06

right and then only two here again two

50:08

long form and then one 2 3 4 five shorts

50:13

right then 1 2 3 so this is the

50:15

situation right now so imagine if we

50:17

were to create a video as a simple um uh

50:21

simple logic here and this is just like

50:23

not even targeted to it. So let's say

50:25

your video if it has to rank you got

50:28

like only two slots here right in the

50:30

top positions only two okay and here you

50:34

got one uh one like 1 2 3 4 five here

50:38

for shorts okay shorts has more odds

50:41

right of somebody coming and clicking on

50:43

it correct two versus five so this is

50:45

the reason why shorts have much more

50:47

reach much more likelihood that a short

50:49

video will get a lot more traffic is

50:51

because of this layout change on the

50:53

homepage Now, in order to get into these

50:56

five, right, these five here, you need

50:59

83% of the people to finish, which is

51:02

why all kinds of dumb videos sometimes

51:03

do better uh than uh otherwise because

51:06

the idea is that the pattern interrupt

51:09

um like top five lists or like say what

51:12

I showed you earlier where you're trying

51:13

to do uh the transitions and the scene

51:15

cuts, right? They're never the same

51:17

person talking throughout, especially

51:19

these days with the attention span being

51:20

low. So, our goal when you're doing a

51:23

short is going to be to land on here in

51:27

this when somebody's coming in. Uh, it's

51:30

going to be here. Okay? Here. If you

51:31

land here, that video will go viral. Uh,

51:34

it's going to get and and it can happen

51:36

even a month after you publish the video

51:37

by the way. Okay? Does that make sense?

51:39

So, this is what we are trying to aim.

51:41

Okay? We want to make our video come up

51:43

here. And if you want to do that, that's

51:46

what I'm saying. You want to make sure

51:47

like in that structure you want to make

51:49

it as exciting as possible, right? You

51:52

want to make it as exciting as possible.

51:54

The story and the transition, the scene

51:56

cuts like so are happening throughout

51:58

the process like so. Okay, does that

52:00

make sense? Hope that part is clear. All

52:01

right. Okay. Now the other uh thing that

52:04

I always try to mention and this is

52:07

quite critical for you to understand is

52:10

that when you try to create videos

52:14

understand what the outcome is with time

52:17

what we're realizing and this is

52:18

something that is very interesting that

52:20

I found out uh with with a lot of

52:22

experimentation we're doing and we also

52:24

know what users are using um and so on

52:26

and so forth right in video express what

52:29

we found out is that for different use

52:33

cases, right? Let's say different

52:37

um um this is not even that easily for

52:40

different use cases, you use different

52:44

structures. Okay, so that means in other

52:46

words um if you go into the workflow I

52:50

gave you right I'll come back here and

52:52

that's what I've been building this

52:53

stuff for you because so much

52:54

experimentation has to be done. So let's

52:56

say here okay we have viral shorts

52:58

workflow we have motion graphics

53:00

animation workflow that we did last week

53:02

then we have idea to 3D documentary

53:04

workflow we have this AI video

53:06

automation which is like storytelling

53:08

copywriting frameworks one right on how

53:10

you can improve there from that

53:11

direction this is like uh this is

53:13

basically first frame last frame

53:14

workflow right and then this is 3D

53:16

animated uh short film and then we come

53:19

back here is realistic why do we have so

53:21

many different unique ones why can't you

53:23

just tell Chad GP hey look create me a

53:25

realistic talking video, I'm good to go.

53:27

Right? Or why can't we just say give me

53:29

a 3D short uh a 3D uh short film or

53:31

whatever and um we or even just say just

53:34

get me a viral video. I don't even know

53:36

what to think. Right? The reason why

53:38

that's the case is because even for AI

53:41

to understand we are trying to give it

53:44

as much internal intricacy details to

53:46

matter. Okay. So for example when we

53:49

come back into video express here and

53:51

this is what no application wants to do

53:53

because they want you to use eat up all

53:55

these credits right they don't care

53:56

about the outcome they want you the

53:58

whole goal right think of it this way

54:00

right there are these platforms out

54:02

there you know there's so many of these

54:03

right if name like uh which all use the

54:06

dance 2.0 to API or Gemini APIs or

54:09

whatever rappers companies selling

54:11

credits and stuff. What's their goal?

54:13

Their goal is for you to fail multiple

54:17

times. So like this scene should not

54:19

make sense. So then you regenerate.

54:20

Regenerate means more credits. So let's

54:22

say it cost $4 for you to create let's

54:26

say I don't know 20 15 second video or

54:28

two clips, right? Let's say two clips

54:33

doesn't matter, right? The goal is to

54:36

for you to come to these two clips, you

54:38

generate a ton of them, okay? 20

54:41

seconds. So then, let's say you generate

54:43

10x of it. Instead of you starting at

54:45

$4, you end up paying $40, okay? And on

54:48

and on cuz at the end you've committed

54:50

already, right? So the outcome is based

54:52

on that. So they don't want to build you

54:53

a workflow which can basically yes

54:56

there's some trial and error but

54:57

eliminates that uh specific recreation

55:00

part. Okay. What we are finding for us

55:03

actually you can create as many as

55:04

possible with video express right

55:06

there's no credits and limits and all

55:07

that mumbo jumbo stuff where other

55:09

people are doing and it's a very hard

55:10

problem that we're solving right

55:12

obviously so then what my goal is is

55:14

okay you want to create a short video so

55:17

then in create with AI right here you

55:21

use the text to video option okay

55:23

because of transitions okay and then

55:26

each transition also should tell a story

55:28

so that workflow right in this

55:31

particular case has to account for that.

55:34

Does that make sense? So meaning each

55:36

use case becomes its own workflow and

55:38

with time it could be dozens of such

55:40

workflows. It could be even hundreds of

55:42

workflows. It all depends on what the

55:44

use case is. Okay? You're trying to do

55:45

an ad, it's a workflow. You're trying to

55:47

create a viral video, it's a workflow.

55:49

You want to create a short, it's a

55:50

workflow. Because behind the scenes,

55:52

these AI tools need lots of examples. It

55:54

needs to be told uh specific things to

55:57

keep in mind. When you're trying to do a

55:58

3D style video, there are certain things

56:00

you got to keep in mind. Consistent

56:02

character videos keep in mind, right?

56:04

There have to be specific things, rules

56:05

of the game, we say that for creativity

56:08

and stuff you got to structure. So what

56:11

happens then with time what we are

56:12

noticing is that those companies which

56:14

are saying okay there's this AGI right

56:17

um which is artificial general

56:19

intelligence or like hey you know you

56:21

can just uh do idea you do and or let's

56:25

say you copy some existing uh video

56:27

whatever and it does the whole sequence

56:30

automatically what's happening here

56:32

let's do the simple math right so let's

56:35

say these uh rapper companies okay which

56:38

all integrate they don't create their

56:39

own video stuff, right? They don't have

56:41

their own servers like we do. So, what

56:42

happens with them? They're basically

56:44

relying on a third party that they're

56:45

reselling, right, back to you for a

56:47

profit on top. So, let's say they create

56:50

like a twominute video even

56:52

automatically and then that cost you per

56:54

video. Let's say per 10 seconds. Even if

56:56

I say a dollar, right? Dollar per 10

56:59

second clip, that's basically 12, right?

57:02

12 uh six uh let's say six

57:07

even at a dollar per 10 second. I've not

57:09

seen a single platform offer that. But

57:10

let's even say right it's going to cost

57:12

you $12 to create this 2-minute video.

57:14

Right now if any one of those sequences

57:18

if any one of the sequences any one of

57:19

these sequences fails like this clip is

57:22

bad let's say right now how do you edit

57:25

it? You're going to have to recreate the

57:27

whole thing. Right? If let's say this

57:28

was bad and then let's say uh this

57:31

sequence was bad, right? Then what

57:33

happens now? They want you to recreate

57:35

the whole thing. And then by the time

57:36

you sit and you finish it, that $12

57:39

becomes $120 and it goes crazy. And this

57:42

is the issue right now because it's all

57:44

about incentives. Okay, it's all about

57:46

like, okay, they charge you by the

57:48

credit, they charge you by the usage.

57:49

What does that mean? Make them use more.

57:54

And this is the reason why what's

57:56

happening in the industry right now is

57:58

that the people who use video express in

58:01

this case and use clone voice or other

58:02

integration to get your own voice and

58:04

all that stuff. What advantage you get

58:06

in the picture is basically where you

58:09

can create as many videos as possible

58:11

but confine that within the workflow.

58:14

Okay, does that make sense now? So when

58:16

the workflow is correct, what the goal

58:18

here will be if you structure it very

58:20

simply with that workflow. What I'm

58:22

trying to achieve here is to make that

58:25

90% accurate, which is going to be the

58:27

same case with these people, but you can

58:30

mix and match here, right? The sequences

58:32

in the bottom. You can delete stuff you

58:34

don't like and all that with that little

58:35

bit of manual work. What that means, in

58:37

other words, is you get the outcome you

58:38

want sooner without all the headaches

58:40

and stuff like that. And people give up.

58:42

Right now, one of the biggest things our

58:44

competition because, you know, we have

58:46

multiple channels. We do a lot of um

58:48

testing and tweaking and stuff like

58:49

that. you know, internally with video

58:51

express, I have access to stuff that

58:53

we've not released yet to you, right?

58:54

And we're testing those things as well

58:56

in real life scenarios. What happens is

58:58

that when we go into the marketplace,

59:00

right, let's say if let's say five

59:03

videos, right? We did it. We did

59:04

everything correct. Let's say five

59:06

videos, we did everything correct. Okay?

59:08

And but it didn't get the results. Now,

59:11

when we go back and we want to

59:12

experiment with five more, let's say, to

59:14

get the outcome we want, that starts

59:17

from scratch. Now, it's like almost like

59:18

a roulette table. You know, you go to

59:20

roulette and you spend $100. It doesn't

59:23

mean that the another $100 you do,

59:25

you're it's going to land on a black or

59:27

uh or red or a specific number, right?

59:29

You start from scratch. So, what's h

59:31

what's happening with all the other

59:32

platforms is they're like, okay, you

59:33

spent let's say $100 on these videos.

59:37

Oh, that doesn't matter. You're going to

59:39

spend another 100red here. I don't care,

59:41

right? It's all based on usage. So

59:42

what's happening is advantage we are

59:44

getting is that when the workflow is set

59:46

up properly okay even if this is true

59:50

okay even if this is true which is

59:51

unlikely it's going to reduce the thing

59:53

with the workflow what's going to happen

59:54

is now the advantage you're going to get

59:56

is that it's not exactly $100 right

59:59

because really it's free right whatever

1:00:02

you got in with video express you can do

1:00:04

even with that $49 plan which is

1:00:06

ridiculous we don't make any money we

1:00:07

lose a lot of money on that even there

1:00:09

right it doesn't mean anything right you

1:00:11

can just go ahead this um 100 here it's

1:00:14

as good as free you understand what I'm

1:00:16

saying now that gives you a massive edge

1:00:18

over the competition that edge is what

1:00:22

will separate people who you know

1:00:24

succeed with AI versus not okay and what

1:00:27

I'm trying to do is bridge that gap as

1:00:29

much as possible with these workflows

1:00:31

such that when we actually do on a

1:00:33

practical basis once we go deep into

1:00:35

this and we're trying to fix this what's

1:00:38

going to happen is in life right like

1:00:40

let's say this is the timeline of

1:00:41

success, right? Sid, let's say. And a

1:00:44

lot of people give up here, here, here,

1:00:47

here. Why? Because it would have cost

1:00:49

them already like $500 in credits right

1:00:51

here. Okay? And they didn't even go to

1:00:54

the next level because iterations and

1:00:56

improvements while you, right, let's say

1:00:58

this is you, you're going to go here and

1:01:00

you're going to keep the timeline

1:01:01

doesn't exist for you anymore, right?

1:01:03

And that's what we're trying to achieve

1:01:04

with V Express. And then you know uh we

1:01:07

get hated on by all these competitors

1:01:09

unfortunately because they all make

1:01:10

their money on their credits and stuff

1:01:12

like that. Easy, right? Integrating like

1:01:14

even we could put seed dance into this

1:01:16

application and say pay $1 for every 10

1:01:18

seconds. Easy stuff. The harder problem

1:01:20

is, you know, we don't want you to pay

1:01:23

all these credits and stuff because it's

1:01:24

going to ruin your creativity. You're

1:01:26

not a Hollywood studio, right? You're

1:01:28

not, let's say, creating an ad campaign

1:01:30

where normally you would pay $15,000 for

1:01:32

a video shoot and it's okay to spend

1:01:34

$2,000 now uh to because it still saves

1:01:37

you money. Correct? So, because of this,

1:01:41

the workflows become very important to

1:01:43

nudge you in the right direction and

1:01:45

also there's a little bit of manual work

1:01:47

to uh fix these to tell the story. Okay?

1:01:50

And all the edge cases along the way

1:01:51

will be optimized based on that. you

1:01:53

don't like anything. Let's say for

1:01:55

example, I don't like this video. I can

1:01:57

crop it, delete it, and move to the

1:01:58

next. Right? I don't like something, I

1:02:00

move to the next and next and next and

1:02:02

so on and so forth. This is the general

1:02:04

idea. Okay? This is what we are trying

1:02:06

to do. This is what we are aiming to do

1:02:08

so that we can uh do as many videos as

1:02:10

we as we want and be able to improve

1:02:12

upon it because the algorithm I tell you

1:02:15

uh is basically uh consistency. Okay?

1:02:19

And this is across all platforms by the

1:02:20

way for shorts. consistency and also

1:02:24

iteration. Okay, so 1% improvement.

1:02:28

Okay. So let's say for example you

1:02:30

learned okay from your analytics right

1:02:32

you learned that uh you know one of the

1:02:34

videos where the uh 83% didn't occur it

1:02:37

was 72% or something right meaning

1:02:40

completion like only 72% of the people

1:02:42

completed for you it looks like amazing

1:02:44

but for a shorts it's bad okay very bad

1:02:47

uh you're not going to get any views so

1:02:49

what happens is then you can come back

1:02:51

and see where did that um uh that

1:02:54

improvement where can it be applied then

1:02:55

you can rework the workflow And then you

1:02:58

can tell like Chad GPD here. Uh let's

1:03:00

say for example I can come back here.

1:03:02

That's why I love these workflows

1:03:03

because it has that context right of

1:03:05

where it messed up. So what I can say

1:03:07

here right is is this one right here.

1:03:09

You see this? I can say something like

1:03:11

this. In the chat window I can say uh

1:03:13

the YouTube short only had 73% retention

1:03:20

um or completion whatever you want to

1:03:21

call it. It can be the both retention

1:03:24

rate. I need it to be uh 83% right. Um

1:03:30

fix this uh fix these prompts

1:03:35

to make sure retention

1:03:38

rate is higher than 83%. Right? So

1:03:42

because of this now it's going to work

1:03:44

on it. It won't be perfect but at least

1:03:47

okay you started not from scratch but

1:03:49

you see this it says for 5 seconds

1:03:51

scenes the biggest mistakes are starting

1:03:52

too slowly. Okay, one idea per scene

1:03:55

instead of new surprise every 1 to two

1:03:56

seconds. See, new surprise every 1 to

1:04:00

two seconds. Weak curiosity and no

1:04:03

emotional payoff at the end. You see? So

1:04:05

now it optimizes it. Shock hook. See

1:04:08

this? Then curiosity. See surprise. What

1:04:13

is a surprise element? See emotional

1:04:14

graph messing it up up and down. Up and

1:04:17

down. Emotional shift. Okay. Reveal. All

1:04:21

right. and then pay off and call to

1:04:23

action. Right? So you see this we are

1:04:26

trying to make it understand that hey

1:04:28

you messed up here. Okay can you improve

1:04:31

upon it here and fix what I did. Now you

1:04:34

get these prompts you can come back into

1:04:37

uh VD express and fix these right here

1:04:39

whichever sequence needs to be updated.

1:04:41

And then you can come back and do this.

1:04:43

Okay. And then what happens is as you

1:04:45

get closer and closer to that 83% you've

1:04:47

got it right. And then what you can do

1:04:49

is you can say let's say for example you

1:04:52

can come back here and say uh this was

1:04:55

perfect and got me 83%

1:04:59

um retention rate.

1:05:01

Can you uh save this as a

1:05:07

um as a uh you know as a memory element

1:05:11

that I can reuse every time the

1:05:14

structure right

1:05:17

okay and it's going to give you advice

1:05:18

on how to go about doing it okay I can't

1:05:21

directly create or edit your persistent

1:05:23

memory but you know you can go and save

1:05:25

it in your memory by saying remember

1:05:27

this preference or something right so

1:05:29

what you can do here is you know when

1:05:30

generating this blah blah blah. You can

1:05:32

now create this and add it into your

1:05:34

memory and and that one might need a

1:05:36

paid account. You cannot probably do it

1:05:37

from free accounts. But you get the

1:05:39

logic, right? You can even say generate

1:05:41

me a system prompt that I can paste as

1:05:45

instructions for future uh to keep this

1:05:48

in mind, right? And it's going to do

1:05:50

something that you can copy paste. You

1:05:52

see this? It's copy paste stuff. So next

1:05:55

time, okay, next time I'll show you what

1:05:57

the trick here is. Next time when you

1:05:59

open up my workflow. Okay, let's copy

1:06:02

this. Um, come here. Build with GPT.

1:06:05

Open up this workflow right here. Okay,

1:06:08

create me a short video on weight loss

1:06:13

journey and how let's say I don't know

1:06:18

um how a good sleep

1:06:22

um got me to lose weight, let's say,

1:06:25

right? Faster than just exercise. So

1:06:27

let's say this is the idea. Um and then

1:06:30

uh follow these instructions, right? You

1:06:32

see what I'm saying? And then just paste

1:06:34

this. This was the uh other uh thing,

1:06:37

right? All this instruction stuff. You

1:06:40

see? And then paste. So what I'm doing

1:06:43

is instructions inside an instruction

1:06:45

based upon the things that I've tested

1:06:47

myself. And they'll be unique to you,

1:06:48

okay? Based on your experimentation and

1:06:51

testing and stuff. And boom, there we

1:06:53

go. It generated this for me. You see?

1:06:55

Um, so now let's just see how this clip

1:06:57

looks like. Okay, let's just test this

1:06:59

stuff because it'll give you an idea of

1:07:01

like um let's go to text to video, go to

1:07:03

advanced, right click and paste. Okay,

1:07:07

and um you can do advanced mode as well.

1:07:09

Up to you. I mean really doesn't really

1:07:10

matter, but let's just click create.

1:07:13

And I'll do it two ways with the without

1:07:15

the enhanced prompt and with the

1:07:17

enhanced prompt. Okay, just for the sake

1:07:19

of it. And uh I'll also do it with

1:07:22

vertical. Okay, create video and then

1:07:25

create video. Now the idea is behind

1:07:27

this I'll show you while this is

1:07:29

generating what you want to pay

1:07:31

attention here is that every time I do

1:07:35

it's going to get intelligent here.

1:07:37

Okay, that means like if I am creating a

1:07:41

video, I made some mistakes, I learned

1:07:43

from that mistake, I can say this was

1:07:45

the mistake, improve it and then that

1:07:47

changes the structure. Correct? And this

1:07:50

after a period of time, let's say after

1:07:52

you've done three or four tests like

1:07:54

this and you got good at the output and

1:07:56

this is what our power users do is that

1:07:58

now you have a system, right? And the

1:08:01

system can be then replicated again and

1:08:03

again. It's not just about like, you

1:08:05

know, people ask us, hey, BJ Express,

1:08:06

can you give us the API or MCP? Hey, we

1:08:09

can do that. We can charge you a lot of

1:08:10

money doing that, too. Um, because at

1:08:12

the end of the day, we got to launch

1:08:13

servers to do these automations. But if

1:08:15

you don't know what you're doing and you

1:08:17

don't have a system in place that is

1:08:19

perfected, no amount of automation will

1:08:22

help you. It's impossible. You're just

1:08:24

going to waste a bunch of credits. And

1:08:25

you were the first one to complain

1:08:26

saying, "Hey, I opened the application.

1:08:28

I created a two-minute video. three of

1:08:30

the scenes of those were completely

1:08:32

junk. The transitions didn't make any

1:08:34

sense and now I had to go pay that extra

1:08:36

$12 or $15 or $20. Now that's not the

1:08:40

future, right? Cost optimization also is

1:08:42

an edge these days. And with electricity

1:08:44

cost, these data centers eating up water

1:08:47

uh you know tons of uh resources sucking

1:08:50

uh all these electric grids and stuff,

1:08:52

it's not going to be cheap. It's only

1:08:53

going to get more expensive. your next

1:08:55

uh variations of your seed dances and

1:08:57

like your Googles and stuff is going to

1:08:59

be more expensive than today. So then

1:09:02

it's not like you're going to just say

1:09:04

okay things are going to get cheaper.

1:09:05

Trust me it's not going to get cheaper.

1:09:07

Okay. So this optimization work matters

1:09:10

and you see this these are the videos

1:09:12

right? So so let me download this and

1:09:15

I'll play this for you. Okay just simply

1:09:17

so you can see these in real time each

1:09:19

one of these and um you can see it why

1:09:22

these are so powerful. Okay. Um, so that

1:09:25

you can understand them as well like in

1:09:27

a nutshell. Um, and and the beauty here

1:09:30

is that the scene transitions and

1:09:32

everything as well are organized and

1:09:34

coordinated, right? So that's the other

1:09:37

thing that we want to basically uh

1:09:39

structure along the way. Okay. So let me

1:09:41

show you this um and play this so you

1:09:43

can see this as an example. I'll just

1:09:44

show you one video uh that was created

1:09:46

in real time. Not everything will be

1:09:47

perfect obviously, but we're trying to

1:09:49

achieve a story that uh makes sense,

1:09:51

right? So let's show you. Let me show

1:09:52

you.

1:09:54

weight faster when I stopped sleeping

1:09:56

less.

1:10:00

I lost weight faster when I stopped

1:10:01

sleeping less.

1:10:06

You see this? What was the idea? Sleep,

1:10:08

right? More sleep helps with weight

1:10:10

loss. So, we're trying to uh drive home

1:10:13

that point and then it it it basically

1:10:16

explained that. Correct? So, that's

1:10:17

basically how we sequence these things.

1:10:19

And then now I can go back into my

1:10:21

screen here and then I can come back and

1:10:23

then again this is the next thing right

1:10:25

fast

1:10:27

what I'm saying and then basically we

1:10:29

come back here and then again we come

1:10:32

back here text to video boom boom boom

1:10:35

come paste here right you can do a

1:10:37

prompt enhancement if you want to I

1:10:39

don't really care it's up to you put

1:10:40

vertical and then boom okay vertical

1:10:43

video could work like this you can put

1:10:44

landscape if you want do that too

1:10:46

doesn't really matter and then those

1:10:48

videos I'll show you as well. So the

1:10:49

cohesiveness I lost weight when I

1:10:52

stopped sleeping less. Okay. And then

1:10:55

it's going to have like that transition

1:10:56

happen uh that goes to the next stage.

1:10:59

And then if you understand the logic

1:11:00

behind why we're doing all this stuff

1:11:02

which we're kind of guiding the

1:11:03

intelligence of a chat GPT here. Again

1:11:05

you can use your free accounts with this

1:11:07

stuff and you can get the workflow for

1:11:08

free. I'm not asking you for any money

1:11:10

and stuff and you're going to improve it

1:11:11

based on your usage. But understand one

1:11:13

of the mistakes is you're going to come

1:11:14

with the automation. Right? The last

1:11:17

thing I want you to do if you don't

1:11:18

follow this video is you're going to

1:11:19

come back here to this page. You're

1:11:21

going to open this GPT, okay? And you're

1:11:24

going to stare at this blank screen, not

1:11:26

knowing what to type here, right? You're

1:11:28

going to stare at this. You're going to

1:11:29

be like, "Blank screen. Where do I

1:11:31

start, right? You could pay $200 a month

1:11:34

to charge GPT, but if you don't know

1:11:35

what to write here, okay, and how to use

1:11:38

it properly, it's all for nothing."

1:11:40

similar to like you could pay $2,000 a

1:11:43

month for like a video tool that might

1:11:45

use something like a seed dance or

1:11:47

something like a Google Gemini Flash or

1:11:50

something. These are the two better ones

1:11:51

these days, let's say for video stuff,

1:11:54

you're going to be spending $2,000 on

1:11:55

it. Same thing. If you don't have a

1:11:57

structure and a system, what is the

1:11:59

outcome you're trying to achieve? And

1:12:01

you go backwards and fill the gaps means

1:12:03

nothing. Okay? And this is the

1:12:05

difference. And uh this is the reason

1:12:07

why even if tomorrow the worst thing I

1:12:09

can do okay and and that's the thing

1:12:11

that I wanted to kind of uh structure

1:12:14

this stuff is you should never be in a

1:12:16

situation where you know costs alone are

1:12:19

stopping you which we don't want to do

1:12:20

and second thing is the structure is

1:12:22

more important and what you're learning

1:12:24

from it reiterating every time is more

1:12:26

important correct so now let's open up

1:12:28

that video as well let's see what's

1:12:30

happening if this thing um generated and

1:12:32

this is the beauty right uh when you

1:12:34

come back into experience this is the

1:12:35

beauty like you can do variations, you

1:12:37

can test them and you can see which one

1:12:39

clicks for what purpose and you don't

1:12:41

have to worry about credits and all that

1:12:42

mama jumbo stuff, right? So there we go.

1:12:45

So let me show you. Okay, that one video

1:12:48

um let's see, let's play this stuff so

1:12:51

you can see it while this is I don't

1:12:53

know why this one failed. I have to

1:12:54

double check. I don't know what happened

1:12:56

there. But let's just try this. Let's

1:12:58

play these two. That's the beauty of

1:12:59

doing multiple videos too, right? So

1:13:01

let's let's see. You can like whichever

1:13:02

one you like, right? One could be an

1:13:04

intro video, one could be an outro,

1:13:05

doesn't really matter. So, let me show

1:13:07

you here. Look at the quality of these.

1:13:09

Okay, vertical first

1:13:14

every day, but the scale barely moved.

1:13:25

Okay, now let's try the other one.

1:13:40

every day, but the scale barely moved.

1:13:51

See the transition it did to make it

1:13:52

exciting basically um for that short

1:13:55

film kind of style, right? So it does

1:13:58

all the thinking in a way. So now you

1:14:00

know what prompts to put, what structure

1:14:02

to put and so on and so forth. Correct?

1:14:04

So this is the idea behind why we build

1:14:06

a system and that system will be

1:14:08

reiterated as time goes on to improve

1:14:10

the workflow that we're trying to do.

1:14:12

Clear? Um

1:14:14

yeah, this is a good question. Let me

1:14:16

answer a couple of questions here as

1:14:17

well. So do I need to uh let me let me

1:14:22

remove this? Yeah. So this is the

1:14:23

question which was like do I understand

1:14:26

correctly that the video format 9x6 16

1:14:28

16 by9 must be specified in the workflow

1:14:31

immediately before building the 16s

1:14:33

after yes good question the reason why

1:14:36

we want to do that is because think of

1:14:38

it like an AI tool right once it gets it

1:14:40

in the prom that it's a 9x6 video in the

1:14:42

beginning it kind of understands that

1:14:45

you know it's a vertical video and it

1:14:46

pulls off of that from its intelligence

1:14:48

same thing with GGP as well if you don't

1:14:50

do that what happens is that you know

1:14:53

you never really see a short video that

1:14:55

is now vertical right maybe an ad like

1:14:58

an ad is different uh but in the

1:15:01

Facebook news feed or something but in

1:15:03

real life right like when you're going

1:15:05

in your phone and you're going through

1:15:06

Instagram reels and uh YouTube shorts

1:15:08

and Tik Tok right you never really see

1:15:10

them as like these uh horizontal videos

1:15:13

so because of that yes you want to make

1:15:16

sure that you specify that in the

1:15:17

beginning okay very important uh now the

1:15:20

other question okay is that um um which

1:15:25

is uh really powerful is that yeah this

1:15:28

can you add a spin video feature that

1:15:29

allows background to spin this can be

1:15:31

done through prompting remember I showed

1:15:33

you this prompting itself if we do a

1:15:35

spin the problem will be where imagine

1:15:37

this right if we gave that spin feature

1:15:40

in that video at what point does that

1:15:42

spin occur is it going to be when

1:15:44

there's a gap or when the character is

1:15:45

speaking when the character is speaking

1:15:46

it won't look good right so that's the

1:15:48

other uh aspect of it now the uh Other

1:15:52

uh thing is basically where which I want

1:15:54

to cover as well is understand that for

1:15:57

shorts if I drive home this point scene

1:16:01

transitions matter more than consistent

1:16:02

character. Okay, especially nowadays

1:16:05

nowadays it's extremely uh hard where

1:16:08

just 1% talking you can uh maintain that

1:16:10

83%. you'll see the best videos will all

1:16:13

be top three, top five videos or you

1:16:16

know what are the five best things to do

1:16:18

or something like that which is more

1:16:19

like what's called a listical where you

1:16:21

keep changing the uh the characters and

1:16:23

the environments and the sequences right

1:16:26

so that's the basic idea of that uh from

1:16:28

that perspective now the other uh thing

1:16:30

as well is that um which is this is a

1:16:34

good point when you are in a workflow do

1:16:36

you ask it to continue your scenes for

1:16:37

example riding the bike in the mountains

1:16:39

and city type thing it all depends

1:16:41

depends on the length of the scene and

1:16:43

what I'm trying to achieve. So for

1:16:44

example, if it is 30 seconds and I have

1:16:47

more room to kind of stretch it, I will

1:16:49

say be consistent and can you stretch

1:16:51

the story and then it'll do it. But

1:16:53

generally speaking what I would think

1:16:57

but actually is that story being

1:16:59

cohesive enough right uh for instance

1:17:01

whatever the the case that I'm trying to

1:17:03

achieve here generally speaking like for

1:17:05

example what I mean by that more real

1:17:07

time I'll tell you let's say the script

1:17:10

was created like this right now you

1:17:12

created some videos now if I say stretch

1:17:15

it to three more scenes is one thing but

1:17:19

if I there's another way to say it where

1:17:21

like hey redo the whole thing where you

1:17:24

fix this thing is much different, right?

1:17:27

Like it will generate the six scenes and

1:17:28

then another two scenes being cohesive.

1:17:30

It's very important about it being that

1:17:32

co u coherent like meaning it's making

1:17:34

sense every sequence. So if I just

1:17:36

stretch it to three more scenes it'll do

1:17:37

it right like here like so but you see

1:17:41

this it kind of doesn't understand the

1:17:43

context of what happened here and kind

1:17:45

of makes a mess of it. See, so that's

1:17:47

why better way to do it is I saw an

1:17:50

issue with scene uh three where um you

1:17:55

know u it is too slow regenerate all the

1:18:00

scene prompts

1:18:02

um to fix the issue then maybe you know

1:18:05

the transition will make more sense

1:18:06

especially if there speech it doesn't

1:18:08

lose it right so it'll generate that you

1:18:09

get the point so I I try to treat it as

1:18:12

such where if I make a mistake and I cut

1:18:15

it like this video is bad and now I try

1:18:17

to regenerate the video. If it still

1:18:19

doesn't come good, then I'll try to

1:18:21

regenerate the whole thing because I

1:18:23

want to make it coherent like I has to

1:18:25

have the consistency throughout. Okay.

1:18:27

Um yeah, keep that in mind. Right. And

1:18:30

um the other thing as well is uh

1:18:35

uh does it work in your own chat GPT?

1:18:37

Yeah, it works with free chat GP

1:18:39

accounts as well as paid ones and stuff

1:18:40

like that. One of the questions is why

1:18:42

did you not do a Gemini and stuff like

1:18:44

that? I keep getting asked quite a bit.

1:18:46

Uh the reason why is uh you know we use

1:18:48

all the tools out there. So in the case

1:18:51

of Google uh they're lagging behind.

1:18:53

Okay. The two players right now in the

1:18:55

game uh are anthropic. Okay. And uh Chad

1:18:59

GPT. These are the two big ones right

1:19:01

now. And Chad GPT in particular their

1:19:05

free plans are really good. Okay. What

1:19:07

you get for it even for the free and

1:19:09

even that $20 a month the cheap uh plans

1:19:12

as well compared to Anthropic. In

1:19:14

entropic the $20 a month plan means

1:19:16

nothing. In Gemini the problem is that

1:19:19

it's not as intelligent. Like if you add

1:19:21

the most latest model with charge which

1:19:23

is 5.6 as of as the time I'm trying to

1:19:25

say is no like Gemini cannot come close

1:19:28

to 5.6. In the case of anthropic their

1:19:29

opus 4.8 is still better than Google

1:19:31

stuff and their fable five model which

1:19:34

released recently is even better than

1:19:36

anything else out there too. Right? So

1:19:38

the problem becomes then I try to tell

1:19:40

you that intelligence is what matters

1:19:42

here and then you go in and you get like

1:19:44

something that is not as good enough

1:19:46

then my message and what you're trying

1:19:49

to do will not match. That's the

1:19:50

problem. You'll try to recreate what I'm

1:19:52

trying to teach you. You'll not be able

1:19:53

to with Google stuff at this point in

1:19:55

time. Hopefully Google changes uh but

1:19:57

they are lagging behind right now. Okay.

1:19:59

Um now and uh the other uh question is

1:20:03

uh um which I wanted wanted to explain

1:20:06

that is uh yeah I did mention with the

1:20:08

beta testers last week. So the issue is

1:20:10

like we already have a bunch of people

1:20:12

that we have and when they drop out we

1:20:14

add more people in. So generally

1:20:16

speaking because we can't handle like

1:20:17

hundreds of people right so if one or

1:20:19

two people like drop out from the beta

1:20:21

group we mention that they'll wait list

1:20:23

of people who we emailed. Okay. So, if

1:20:25

we didn't receive a response, probably

1:20:26

you're in that wait list, but uh you

1:20:28

know, just watch out. We will uh add you

1:20:30

in later on because what we need is

1:20:31

people who are in our like uh our chat

1:20:34

groups and stuff to uh coordinate with

1:20:35

us. Okay. And uh now the other uh thing

1:20:40

as well that I wanted to do is uh um

1:20:44

which which is very critical question

1:20:46

that comes in quite a bit is basically

1:20:49

like this question right here. the voice

1:20:52

over stuff you can do it for narration

1:20:54

videos okay in um video express but I

1:20:58

want you to understand right think of it

1:20:59

this way when you come into the

1:21:01

application and you try to do like

1:21:04

videos which are transitioning to

1:21:05

different things a lot of people

1:21:07

struggle when let's say you upload one

1:21:09

voice what what happens to this person's

1:21:11

voice right you're not going to be able

1:21:13

to control it especially once we're

1:21:15

going into more of these structures for

1:21:16

workflows and stuff like that if you

1:21:18

give only one voice it's going to the AI

1:21:19

is going to expect five voices do you

1:21:21

know five voices will be which are the

1:21:23

specific five voices. It's going to

1:21:24

waste a lot of your time to make those

1:21:25

decisions. So the way we are going to do

1:21:27

it in the right now the way we are doing

1:21:30

it is you can come to any video right

1:21:32

click and you see voice changer you can

1:21:34

come here and just upload your voice

1:21:37

okay as long as the accent is going to

1:21:39

be what you want let's say UK accent or

1:21:41

Australian accent stuff and the male

1:21:43

voice or whatever it's going to make

1:21:45

sure that the voice is similar to the

1:21:47

reference like the voice you're giving.

1:21:48

Okay, this is the best option especially

1:21:50

when it's so many fast-paced edits

1:21:52

coming like every four or 5 seconds or

1:21:54

every 10 seconds or something you're

1:21:55

going to keep changing the voice. How

1:21:56

are you going to ensure that you know if

1:21:58

uh it moves from one person to the other

1:22:00

and stuff? How do you maintain that

1:22:01

consistency and also more importantly

1:22:03

how do you make it exciting still? So

1:22:05

the voice changer is a much more easier

1:22:07

feature and you can apply it against um

1:22:09

the whole timeline as well very easily

1:22:11

uh is the reason why we did it. Of

1:22:13

course there are better ways to do it

1:22:14

improvements and stuff. We are working

1:22:15

on that behind the scenes based on our

1:22:17

testing and stuff. But uh when you have

1:22:19

too many people talking in a in a video,

1:22:22

this is the best option right now. Okay.

1:22:24

To ensure that that consistency remains

1:22:26

throughout. Okay. And uh the other uh

1:22:29

question that came in as well that I

1:22:32

want to uh ensure that you understand is

1:22:35

this okay is basically the workflow

1:22:40

stuff. We're never going to charge you

1:22:41

anything. So when you go into this link

1:22:43

and then you come into the workflows tab

1:22:46

right now what's happening is you have

1:22:48

this right so let's say you go into one

1:22:50

of these applications which compete

1:22:52

against us let's say they are like $99 a

1:22:54

month and on even with credits and stuff

1:22:57

then that's not enough there are these

1:22:59

course creators and these cool

1:23:01

communities where they'll say I created

1:23:03

a workflow for this application right

1:23:05

and then they'll say buy that for

1:23:07

another how muchever they feel like is

1:23:10

correct right they'll say for the

1:23:12

workflows pay me another $100 a month or

1:23:15

something like that for the training and

1:23:16

stuff. Now, this is where the f the

1:23:20

whole world is headed towards. What I

1:23:22

want to tell you is this stuff is free.

1:23:24

It's always going to be free. These

1:23:26

workflows, it's going to take me hours

1:23:27

and hours of my time and thousands of

1:23:29

hours of our team to test it and stuff,

1:23:30

but it's going to be for free. Why?

1:23:32

Because at the end of the day, you don't

1:23:34

if you save money here, right? And don't

1:23:37

have to pay monthly fees here, then at

1:23:39

least you'll succeed with our

1:23:40

applications with uh without restricting

1:23:42

yourself into what you can do or what

1:23:44

you cannot do, right? And um often times

1:23:46

what happens with people, they'll come,

1:23:48

they'll say, "Wow, that's amazing. That

1:23:49

course, it's going to give me all the

1:23:50

workflows. Let me go jump in and it's

1:23:52

going to be $100 a month." Then they'll

1:23:53

have upsells inside and all that stuff.

1:23:55

And then they're going to say, "Oh, by

1:23:56

the way, you're going to need your uh

1:23:58

applications now." Okay, this is one of

1:24:00

the applications you can use. This is

1:24:02

another one up to you. All of them come

1:24:04

with usage fees, credits, and stuff like

1:24:05

that. And then you jump into that.

1:24:07

Before you know it, you've already spent

1:24:08

$1,000, right? And you haven't even

1:24:10

begun. And people give up by that point

1:24:12

because they feel like they've not

1:24:13

gotten anything for that investment that

1:24:16

they put in. And I'm saying even if you

1:24:18

merge all of our applications together,

1:24:20

it is not even thousand bucks,

1:24:24

so on and so forth. And imagine that,

1:24:27

right? how much of a leverage and edge

1:24:29

you have over your competition because

1:24:31

of that because we constantly keep

1:24:32

updating our applications and stuff. So

1:24:34

coming back to the point these workflows

1:24:36

will be free and have more of them

1:24:38

planned out but understand that it's not

1:24:40

going to be a beall and endall. The goal

1:24:41

will be to improve upon these and build

1:24:44

upon what I built for you. Okay, that's

1:24:46

the goal because this is a starting

1:24:48

point. Everybody's use case will be

1:24:49

different. Somebody who's trying to

1:24:51

create a product based video. Somebody's

1:24:52

trying to sell a service. Somebody's

1:24:54

trying to create a viral video just for

1:24:56

monetization for YouTube. Somebody might

1:24:58

be able might be wanting to create just

1:25:00

to test their feet to get what's going

1:25:01

to get them views uh to get some

1:25:03

motivation to continue different

1:25:05

objectives. Okay. And each of those

1:25:08

objectives will have an own uh its own

1:25:09

workflow. Okay. And that is my job to

1:25:12

tell you which workflow is good for what

1:25:14

and then what do we bring it back into

1:25:16

video express. Okay. That's the general

1:25:18

idea of what we are trying to do and

1:25:20

these will be for free. You'll not be

1:25:21

charged. So instead of spending

1:25:22

thousands of dollars on courses like if

1:25:24

you want me to add any workflows, just

1:25:26

let us know. Okay, this is the email.

1:25:28

Say Paul, I want this workflow. My team

1:25:29

and I were on top of this stuff. We're

1:25:31

going to be building you. And if that

1:25:32

feature not exactly clearly available in

1:25:35

video express, let's say we're going to

1:25:37

add them in as well. Okay. Um that might

1:25:39

be transition someone was asking that

1:25:41

could be for example uh you know fast

1:25:43

cuts. You see these more uh styles here

1:25:46

like we have whatever it might be. Okay.

1:25:48

The end goal is you still need that

1:25:50

manual editing. I'm going to tell you

1:25:52

for a fact no matter what these

1:25:53

so-called faceless channel gurus will

1:25:55

tell you um YouTube and Tik Tok and all

1:25:58

these people, they do not want AI slop.

1:26:02

Okay? They do not want it. And how do

1:26:04

you prevent it? You're still going to

1:26:06

put that 3% or so creativity. Okay? To

1:26:10

make it unique. Okay? And that is what

1:26:12

is going to separate you from everybody

1:26:14

else. Because the other ways are

1:26:16

climbing up the hill where a channel

1:26:18

gets demonetized or a channel does not

1:26:20

get uh much traction going forward. Then

1:26:22

what happens? Then you're going to give

1:26:24

up cuz you're trying to climb up the

1:26:25

mountain, right? It's not fun trying to

1:26:27

create more YouTube accounts or more Tik

1:26:28

Tok accounts or whatever cuz you did not

1:26:31

get success from one. It's much harder.

1:26:32

Do it right the first way. you add the

1:26:34

3% creativity which you're going to

1:26:36

simplify

1:26:39

the other people who'll have like a

1:26:41

graph like this, right? Like they're

1:26:43

going to have it's so funny and I've

1:26:45

seen it so many times. It's going to be

1:26:46

like this, right? Their graph is going

1:26:48

to come like this and crash like this

1:26:50

because this peak right here was not

1:26:52

organic. Okay? It was done through

1:26:54

manipulation of the algorithm pushing it

1:26:56

to the limits and all that kind of stuff

1:26:57

and then it'll drop automatically.

1:27:02

What I try to tell you is it's better

1:27:04

just being slow and ups and downs, up

1:27:07

and ups and downs, ups and downs. And

1:27:09

you have that graph like this. Let's say

1:27:10

this is the graph. Much better because

1:27:13

you'll have a more consistent flow. And

1:27:15

here, every time you try to start and

1:27:17

you keep starting now, every time you do

1:27:18

it, you give up. It's too much. It's

1:27:20

really, really, really hard to catch up

1:27:21

after, right? And this is what's

1:27:23

happening with all these people talking

1:27:25

about it, which they miss out on. and

1:27:26

they're going to tell you, "Hey, I used

1:27:28

AI to generate the whole video for me."

1:27:30

But they don't they miss the point where

1:27:32

if you did do that, it's not going to

1:27:34

get results because uh these algorithms

1:27:36

and these platforms are smarter and

1:27:38

smarter with time. You think Facebook a

1:27:39

billion trillion dollar company uh you

1:27:41

know Instagram or let's say YouTube,

1:27:44

right? You think they're like crazy

1:27:45

where they're so silly that you can just

1:27:47

automate your way. Not so much, right?

1:27:48

So these little changes like editing it

1:27:50

in here, removing cropping here, here,

1:27:52

here. These kind of elements matter a

1:27:54

lot which I taught you today. Okay. And

1:27:56

then with time, it's going to become

1:27:57

even more important once you know how to

1:27:59

uh build the system for you to reuse and

1:28:01

stuff like that. Okay. And we have a lot

1:28:03

of other cool updates coming up in Video

1:28:04

Express. Watch out for that. And of

1:28:06

course, we're going to be having updates

1:28:08

with uh CloneVoice V4 coming up. We've

1:28:11

been working extremely hard on that with

1:28:13

the integration with Video Express as

1:28:15

well. So, uh if I write it here, Clone

1:28:17

Wars uh V4 that should be coming

1:28:19

extremely soon, probably within days.

1:28:22

Okay. I'm hoping by next week, but you

1:28:24

know, if something changes, a better

1:28:25

test might change. It's going to be V4.

1:28:27

Uh it's going to have 400 languages

1:28:30

uh that we support 400 plus for cloning

1:28:32

and generation and stuff. And on top of

1:28:34

that, we improved uh better um you know

1:28:37

uh text to speech generation for your

1:28:39

clone voices and stuff in certain

1:28:40

languages like Hindi or Thai or um some

1:28:43

of these languages out there which we do

1:28:45

know can be improved upon. So that will

1:28:47

be coming next week. And after that,

1:28:49

we're going to be working on artistly

1:28:50

v7. It's a huge project that one uh

1:28:53

because we have to update every single

1:28:54

little feature uh from the ground up. So

1:28:57

that's coming up as well right after

1:28:58

that. And then we have our next uh video

1:29:01

express update in between that which is

1:29:03

going to be more on optimization and uh

1:29:05

certain things as well. So watch out for

1:29:06

that like that first frame last frame

1:29:08

will allow you to upload your own images

1:29:09

for that and um you know improve that

1:29:12

and stuff like that. So right now for

1:29:14

example in video express uh when you go

1:29:16

into create from AI in first frame last

1:29:18

frame uh you see in my account there is

1:29:21

this upload it says beta you're going to

1:29:23

get this feature in your account. It

1:29:24

won't be in your account right now. Uh

1:29:26

we're still doing some tests on this

1:29:27

stuff to improve it and simplify the

1:29:29

stuff. So stuff like this right we're

1:29:32

going to release it to you and announce

1:29:33

it. Won't be a major announcement

1:29:34

because it's a pretty straightforward

1:29:36

feature but watch out for our Facebook

1:29:38

group and stuff for these updates. Okay

1:29:39

cuz I'm testing it internally. We want

1:29:41

to make sure we test everything

1:29:42

throughout and then when we're satisfied

1:29:44

with the test, we want to release it

1:29:45

rather than rushing it. Okay. So, that's

1:29:47

the basic idea. All right. Okay. So, uh

1:29:50

hope that uh helps and uh yeah, this is

1:29:53

the good point that V4 we are hoping

1:29:56

will fix that. We've tested Ukrainian in

1:29:58

that V4 specifically. I can tell you it

1:30:00

was way better in the V4 of Clone Wars

1:30:02

than uh V uh V3 the existing version.

1:30:05

One thing I do want to mention though is

1:30:07

uh what we found out with Ukrainian and

1:30:10

some of these languages is that some

1:30:13

words the pronunciation might be

1:30:14

slightly off. Okay, in that case what I

1:30:17

what we suggest is that uh you know like

1:30:19

how you know for example in English we

1:30:21

stretch certain words with exclamation

1:30:23

marks or uh you know certain ways we can

1:30:25

pronounce those words right we change

1:30:27

that way we write it just keep that in

1:30:29

mind uh it's going to happen only like

1:30:31

2% of the time or so u because of

1:30:33

whatever AI like that we'll I'll explain

1:30:35

that in the tutorials okay so meaning

1:30:37

it's be substantially better than it is

1:30:39

right now the clone wars v4 for these

1:30:41

languages like these and we are the only

1:30:42

ones working on these too you know other

1:30:44

platforms have given up multilingual

1:30:46

long ago for voice cloning and stuff

1:30:48

like that. But what I do want you to do

1:30:50

is it'll be 99% perfect. That 1% might

1:30:53

be based upon what you type. You want to

1:30:55

adjust it. Is that clear? That's what I

1:30:57

would suggest you to do. Okay. Um

1:30:59

basically okay. So um all right, that's

1:31:02

pretty much it guys. So and gals who are

1:31:04

here. I appreciate you all. So that um

1:31:06

you know I know there were some

1:31:08

technical issues in the start. So I

1:31:09

appreciate you all doing it. I'll try to

1:31:11

edit it to chop it out so you can watch

1:31:12

the replay easier. Uh but understand

1:31:14

that we're working on all these things

1:31:16

and uh also I'll be updating these

1:31:18

workflows with time. They'll be 100% for

1:31:20

free as always. Just make sure that when

1:31:22

you're coming into video express don't

1:31:24

underestimate the text to video feature,

1:31:27

okay? It's amazing. And more

1:31:30

importantly, don't underestimate the

1:31:32

fact that it doesn't have consistent

1:31:33

character in text to video. Make sure

1:31:35

that for shorts and certain specific use

1:31:37

cases where you know you need the scene

1:31:39

cuts and stuff, nature videos, you know,

1:31:42

videos where like, you know, you're

1:31:43

trying to tell a story that is uh moving

1:31:45

in different um directions and stuff and

1:31:47

there shorts like this scene cuts, best

1:31:49

of five lists, top five, you know,

1:31:52

listicles, we like to call them, list of

1:31:54

five things you should do when you go to

1:31:56

Wyoming and uh travel videos, you know,

1:31:59

stuff like that. Uh amazing. Okay. and

1:32:02

it's probably among the very best in the

1:32:05

uh world right now regardless of the

1:32:06

cost. Okay, even if you pay $4 for 10

1:32:08

seconds, it's going to be the same as

1:32:09

what you get here. And it has specific

1:32:11

use cases. Understand that. And that's

1:32:13

both for landscape. I'd use landscape

1:32:15

mainly. Uh this landscape video uh for

1:32:18

ads, right? Uh Facebook ads, YouTube

1:32:20

ads, that kind of stuff. And then

1:32:22

vertical video is for shorts. Okay, just

1:32:26

understand that. Very important. Then

1:32:27

once you get that, you can follow the

1:32:29

workflow I gave you today. experiment

1:32:30

with it and you'll do really really well

1:32:32

um from what you are doing and of course

1:32:34

there's no credits monthly fees and

1:32:36

stuff so you have nothing to worry about

1:32:37

in terms of experimentation let us pay

1:32:39

all the electricity costes GPU server

1:32:41

cost there's millions and millions of

1:32:42

dollars we're spending uh but at the end

1:32:45

of the day our goal is to give back to

1:32:46

you because you've been loyal to us and

1:32:48

a lot of the times you know a lot of you

1:32:50

folks buy other stuff from us like clone

1:32:52

voice and stuff we don't have to spend

1:32:53

advertising u like all these other

1:32:55

platforms do which are just a bunch of

1:32:58

rappers around these platforms from

1:33:00

Google or bite dance seed uh seed dance

1:33:02

and stuff like that and then they mark

1:33:04

it up on top. So then they basically

1:33:07

will squeeze you dry. Squeeze you dry in

1:33:09

parallel generation. Squeeze you dry in

1:33:11

terms of charging you for workflows.

1:33:13

Squeeze you dry in terms of the features

1:33:15

throttling all this stuff. We don't do

1:33:17

any of that stuff. But what we do want

1:33:19

you to understand is that don't

1:33:21

underestimate the true power of this uh

1:33:23

what we have with express and it's only

1:33:24

going to get better better with time. So

1:33:26

what you have today can do amazing

1:33:28

videos. And the sad part is we could

1:33:31

charge 10 times more for Video Express

1:33:33

and make more money. Uh and sometimes we

1:33:35

get uh some people to say like, "Hey,

1:33:37

I'm a serious agency. I'm only paying

1:33:39

like what is it about a couple of

1:33:41

hundred bucks or something, right, for

1:33:42

radio express, the unlimited license,

1:33:44

$200, it's too cheap. I don't trust it."

1:33:46

And we hear that and it hurts us because

1:33:49

what we are trying to do is help

1:33:50

everyday people like you um to actually

1:33:52

use it and then it backfires against us.

1:33:55

And that's why what we plan to do in the

1:33:56

future is we're going to do an

1:33:57

enterprise plan. Uh it's going to have

1:33:59

specific features like single sign on.

1:34:02

Um it's going to have like uh some other

1:34:04

specific features where we're going to

1:34:06

charge us money. They're going to charge

1:34:07

extra money on it just like these other

1:34:09

competitors of ours are doing. Uh you

1:34:11

know who they are. They'll claim free

1:34:13

usage of uh seed dance for like a day or

1:34:15

two days and never works. Uh and then

1:34:18

you jump into they rebuild you and all

1:34:20

this other stuff. Shady shady characters

1:34:21

unfortunately. Um but our our thought

1:34:24

process is like we have to go into that

1:34:26

direction like APIs and all that stuff

1:34:29

we're going to for the enterprise heavy

1:34:31

uh agencies and stuff like that we will

1:34:32

do that but for you it'll always be this

1:34:34

you'll be grandfathered in with the

1:34:36

updates and stuff like that regardless

1:34:37

of our cost and our margins are super

1:34:39

thin but it doesn't matter because uh

1:34:42

word of mouth and stuff like that is

1:34:44

helping us grow right now uh just for

1:34:46

reference last year alone more than

1:34:49

100,000 people purchased our products

1:34:51

it's not a small amount right?

1:34:53

Ridiculously high. So we are the few

1:34:55

people in the planet who can say that.

1:34:58

Okay. Uh so that means we're learning a

1:35:00

lot from all the people around the world

1:35:02

and stuff and that will give us an edge

1:35:04

because a lot of people give up like we

1:35:05

had the Ukrainian language, people give

1:35:07

up in Bangla language, people give up on

1:35:10

some of these uh languages and Nepali

1:35:12

language because it's too small of a

1:35:13

market, right? European and English is

1:35:15

good enough. We want for the whole world

1:35:18

to do it. That's the only way it's going

1:35:19

to work is through word of mouth and

1:35:20

improving it through these platforms.

1:35:22

Correct. So anyway, that's the point

1:35:24

here. Just understand that these

1:35:25

features they look like a little bit

1:35:26

complicated and stuff, but they're are

1:35:28

usage based, use case based. Okay? And

1:35:30

we want to work towards that. Okay? So

1:35:32

have a great day on navigator in the

1:35:34

world. Um I wish you all the best. We'll

1:35:35

look forward to seeing you on our next

1:35:37

training. More workflows will be shared

1:35:39

and other stuff as well in our Facebook

1:35:40

group and other places. Okay? So watch

1:35:42

out for that.

More transcripts

Explore other videos transcribed with YouTLDR.

Get the TLDR of any YouTube video

Transcribe, summarize, and repurpose videos in 125+ languages — free, no signup required.

Try YouTLDR Free