YT YouTube2Text

YouTube2Text → Transcripts → My AI UGC looks 100% real because I don't use Seedance (copy me)

My AI UGC looks 100% real because I don't use Seedance (copy me) — Transcript

by Alex Fedotoff · 4,923 words · 695 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00In the next 15 minutes, you'll learn how
0:01to create realistic UGC content like
0:04this that can sell your product for
0:06hundreds of thousands of dollars or even
0:08millions of dollars. In the last 30
0:09days, I made over $1 million with one of
0:11my e-commerce stores, and this is one of
0:14the formats that's making us the most
0:15money. For the purposes of this video,
0:17I'll use a regular product like this one
0:19on screen as the something that doesn't
0:21have anything special. It's not unique.
0:24Anyone can source this product and sell
0:25it, and I'll show you how to take this
0:27product and package it through the ads
0:29into something that people buy millions
0:31of dollars a month. And by the way, this
0:33store generates 3 million visitors a
0:35month, and they have 850 active ads
0:37right now. To make a UGC video with a
0:40product, we need an actual photo of the
0:42product itself. And it's very important
0:44to have a real person in this ad. So,
0:46we're going on Pinterest, and we're
0:47looking for pictures of a woman that
0:49looks very realistic and high-quality
0:51picture. And then we're just saving one
0:53of these pictures. There's another photo
0:55I found here. It looks realistic, but
0:57when you zoom it in, it's kind of like
0:59blurry, and it's not as detailed, so we
1:01don't want to use this one. This will
1:02seriously impact the quality of the
1:04video because the quality of the footage
1:06will not be as good and as detailed. For
1:09generating the images, we'll use
1:10ChatGPT. We'll just upload the image
1:12that we downloaded from Pinterest here,
1:14and then we'll just press it from here.
1:16I want you to analyze this image and
1:18create something similar, uh like a lady
1:20that's uh holding the microphone in her
1:23hand. Uh so, it looks very realistic.
1:25She just sits in a car and is trying to
1:29kind of like record some video or
1:30something in a selfie format. Create the
1:33highest [snorts]
1:34detail uh photo possible here. And the
1:36reason I just said this lady has to hold
1:38the microphone in her hand is because,
1:40number one, it looks more authentic. If
1:42someone sees it, it looks like a real
1:44person just recording a video. And
1:46number two is because some people might
1:48be asking in the comments why the video
1:51uh or audio quality is so crisp, and
1:54there's an easy the for that because she
1:56just literally holds the microphone in
1:58her hand. So, now it's working to create
2:00an image for me and we'll create another
2:01chat on ChatGPT here in the meantime.
2:03And in this one, we will upload the
2:05picture of the product. In this case,
2:07it's this astaxanthin product. And then
2:09I'm recording a prompt like this. I want
2:11you to find the me images for this
2:13product from different angles in a very
2:15high quality. Also, the images of the
2:19capsules themselves. Show me visually
2:21how big they are. So, I have a good
2:24understanding of that and I can use
2:26those pictures to create realistic
2:29content or put it on my website. I just
2:31need like highly detailed images from
2:34different angles of this packaging of
2:36this particular product. So, the reason
2:38we're generating this right now is
2:39because we want to have images of the
2:41product from different angles. So, when
2:43AI creates me the footage, it's as
2:46realistic as possible from different
2:48angles. It has a full understanding of
2:50the dimensions, how big it is, how it
2:52looks from different angle, and has the
2:54full picture. All right. So, now we'll
2:55switch to another chat and we'll see
2:57what we got. Oh, we got this lady here.
2:59So, she is very realistic. So, if we'll
3:03just like zoom it in, I can't even zoom
3:05it in like but she looks pretty
3:06realistic. Now, the reason we took the
3:08picture of this lady in the car is
3:10because if we would just like take like
3:12say to AI, "Hey, create an image of a
3:15person in the car." Then it could be
3:17blurry on the back end. It could be not
3:19as crisp. It would be some like little
3:21like fuck-ups and we don't want that to
3:23happen. So, we just want to give AI a
3:25reference that is as close to real as
3:28possible and then it creates something
3:30that is very realistic as well. So, if
3:32we compare these two, this is the
3:33original one that was there and this is
3:36the one it created. You can see very
3:38realistic. Now, she holds the microphone
3:39in her hand. And everything from the
3:41hair here to the skin to little like
3:44imperfections on the skin to the
3:46microphone to the cars on the background
3:49to the things even like the back seats
3:51and shadows there. Everything is hyper
3:53detailed and very like, I mean,
3:56realistic. The most important thing when
3:58you're creating AI content is the
4:00reference. The more realistic it is, the
4:02better quality AI content you will get.
4:05Now, look, I [ __ ] up because I said to
4:07create a realistic content to put it on
4:09my website. That's why I generated this
4:10kind of footage that is mostly to put on
4:13the website, but it's not necessarily
4:15the product itself. So, I'll have to
4:17regenerate this one. All right, so I'll
4:18recreate clean shots of the pouch and
4:20soft gels front angle back with no added
4:23captions or graphics. Boom. It's working
4:24on it. So, this lady will be talking in
4:26the video, but the important part is
4:28what she'll be talking about, the actual
4:29script of the video. And to create a
4:31script that gets someone to pay you
4:33money, to pay for your product. And in
4:35order to do that, you have to use one of
4:37the angles or micro angles for the
4:39product that you're selling. So, in my
4:40Clod account, I pulled all of the
4:42Clarity MD ads that they're running
4:44right now, all of the angles, all of the
4:46micro angles they're running in order to
4:48identify the best angle to build my ad
4:51around. So, I'll download this image
4:53from here that we generated. So, now I
4:55just pasted the image of this lady here
4:57and I'm recording this prompt. Hey, so I
4:58have the image of this red-haired lady
5:00sitting in a car. From the research that
5:02we did about Clarity MD, I want you to
5:04choose a micro angle that would best fit
5:07her, her age, how she looks like, the
5:10background where she's sitting. She's
5:12sitting in a car. Choose that micro
5:14angle and let me know here which one you
5:16would choose. And now I'm just sending
5:18it. Let's see what it does for me. All
5:20right, so it selected the angle for me.
5:21I'm returning this. It's kind of like
5:23reverse hook. Yeah, I'm definitely
5:24returning this [ __ ] because not because
5:26the most it's the most important
5:28astaxanthin for kittens in the United
5:30States right now. And it's not because
5:32the soft gels are oil-based as it should
5:34be so your kittens can actually absorb
5:35it, right? This is like very good one
5:37that's kind of like reverse selling.
5:39Now, it generated some of the images for
5:40me. It's actually [ __ ] up with one of
5:42the images because it opens on the side
5:45instead of the top, but it
5:46self-corrected. You can see the AI just
5:49realized that it [ __ ] up and it
5:51regenerates that version again because
5:52it has to open here, not here. Okay,
5:55boom. Now it generated the right
5:56version. Okay, cool. So I have all of
5:59the images here.
6:01I'll just open each one of them and I'll
6:02save them. So now I'll just download all
6:05of the images at once. Now one thing
6:06that I found here is that this lady here
6:09is not ideal kind of like age persona
6:12for this brand, but it suggests me to
6:14write a specific angle for younger
6:17audience for this particular product.
6:19And by the way, the prompt that I used
6:20to do this whole research for the brand,
6:22you can do this for your brand, you can
6:24do this for your competitors' brands,
6:25direct or indirect. I'll leave this in
6:27the description so you can use it
6:29yourself. So even though this brand is
6:30not using younger people for ads or not
6:32targeting that demographic, it suggests
6:35me the angle of skin and UV protection
6:37that nobody is using in that niche. So
6:40that's what we will write the script
6:41about. I'll write the script
6:42specifically for UV protection because I
6:45think that angle could be massive. Now
6:47as I'm looking at it, this [ __ ] is
6:48crazy. I've been in e-commerce for 10
6:51years. And before you know, let's say we
6:53want to sell a product like that. We
6:54have to get a product. We have to get it
6:56to our office or we have to get it to
6:58the photo shoot studio. We have to like
7:00set up the lighting. We have to bring in
7:02the photographer. He has to make those
7:05photos. Then he has to clean them up or
7:07like clean them up in Photoshop. Now I
7:09did all of this [ __ ] here in like 2
7:11minutes. The highest quality images you
7:13can find. This is insane. Now I
7:16personally have $200 charge GPT account,
7:18but you can generate this for free
7:20probably with free account. All right,
7:21so I have a script now. She taps on the
7:23side window. This window is aging one
7:26side of your face faster than the other.
7:28Actually pretty good angle here. Your
7:30windshield block blocks the most UVA.
7:33Yes, so yes, sunscreen every day
7:34non-negotiable. And bum bum bum. So this
7:37is placebo-controlled trial out of
7:38Japan. As the scientist, she introduces
7:40the product pretty fast. So like
7:421-minute video that introduces people to
7:44the product and builds on this new angle
7:47that was not even used on this brand
7:49that like if we crack it, we basically
7:51open up the whole audience that no one
7:53else is tapping into. Now, there's
7:55another one that's pretty good. I'm
7:56redhead, I burn in about 10 minutes.
7:59This is great one. I'll I'll actually
8:00use this one. I like it a lot because
8:02she is redhead and now like when redhead
8:05people see her ads, they'll be
8:07resonating with her big time. Now, the
8:08next step here as I have the script
8:10broken down, I'll need to generate the
8:13prompts for each one of these segments
8:15that I can then take to ChatGPT to
8:17generate an image for each one of these
8:20segments. Otherwise, if I'll just use
8:21the original photo here, it will look
8:23like [ __ ] So, generate me prompts for
8:26each one of these segments that I can
8:28use uh for some image generation AI to
8:33generate super detailed images, but
8:35based on the script, you can see they're
8:37all looking a little bit different. So,
8:40um generate me the prompts that I can uh
8:42make the images with. Only now we'll
8:44take these prompts and we'll be
8:45generating images in ChatGPT that we'll
8:48then use in our AI video creation
8:50platforms. Otherwise, if you skip this
8:52step, you'll get AI slop that does not
8:54look realistic that people will not buy
8:56from. Okay, so I'm copying the first one
8:58and pasting it back in here. In this
9:00shot, I'll insert the image that I
9:03generated plus the prompt from Claude
9:05and I'll I'll have to give it some extra
9:07instructions. Hey, so this is the
9:09original image. We want the same lady to
9:12look the same with all of the images
9:15that we'll generate, but with specific
9:16instructions that I have provided here.
9:18Now, this was the original image that we
9:20have created. Look at this one.
9:22Hyperrealistic, like even the light, the
9:25shadows, I mean the the roof now it's
9:28more visible here, very realistic. So,
9:31I'll just continue with other prompts
9:32now. So, we generated the next one. Now,
9:34you can see she's tapping her window, so
9:36it looks very realistic. It look you can
9:38see the still the light, the shadows,
9:41like everything is hyperrealistic. But
9:43only thing she's not holding the mic
9:46in her hand. Now it's just attached to
9:48her clothing. So these are little
9:49details, but she might have just like
9:51attached the mic so she can knock on the
9:53window. And you can easily regenerate
9:55the image from the start where she
9:57doesn't have to hold the mic in her
9:58hand, just has it here. So now we'll use
10:01this one as a reference for the next
10:02one. Now you can see generated me the
10:04image where she's just holding the
10:06random sunscreen with no labeling, no
10:08branding, and so we'll have to
10:09regenerate this image with a specific
10:12brand name so it looks more realistic.
10:14So we'll just use this one for the
10:15purposes of this video. Hey, I want you
10:17to replace the sunscreen that she's
10:19holding because it has no branding with
10:21this specific brand of sunscreen and its
10:24packaging. Now for comparison, I'll show
10:26you Nano Banana Pro. This one is paid uh
10:29compared to ChatGPT, and I'll show you
10:31difference in quality between
10:32generations of Nano Banana and ChatGPT.
10:35So now it regenerated the image with an
10:37actual sunscreen. Now it's a lot more
10:39realistic. Now in Nano Banana, I
10:41uploaded the same prompt. I uploaded a
10:43photo of the lady and also the photo of
10:45the sunscreen, and now we'll compare the
10:47quality in output. I also selected 4K
10:49quality and 9 by 16 aspect ratio. So
10:52this one will cost me 24 credits. We'll
10:55see the output.
10:56All right, look at this. It's [ __ ] it
10:58up. Like this was the original one, and
11:01now this is the lady. It doesn't look
11:03like the same one. So I would not use
11:05Nano Banana for this. If you're using
11:07ChatGPT, just stay there, use the same
11:10model, and it's better quality and
11:12cheaper. Now we can clearly compare the
11:14quality. Even though I gave Nano Banana
11:16clear instructions, "Hey, use this
11:19[ __ ] product inside of the photo." It
11:22didn't for some reason. Just I probably
11:25had to prompt it again. Just didn't get
11:27what I wanted. And so now this one is
11:29extremely clear. It's using the right
11:32product. Even the lady doesn't look very
11:34similar, even though I uploaded the
11:36exact reference image for it to use. So,
11:40Nana Banana in this case sucks. Now,
11:42there's some little details that are
11:43better on Nana Banana. You can see the
11:45details like little like things on the
11:48window, like the shadows, like
11:49everything is very precise. ChatGPT is
11:52not as detailed, but you can still
11:54refine it later and it's not a big deal.
11:57One part of our script is b-roll insert,
11:59which means it's not a image of a person
12:02of a lady. In this case, it's just a
12:04b-roll insert of a like round cells and
12:06stuff like that. So, for this one we'll
12:09use an actual product image that we have
12:12generated of the soft gels themselves.
12:15So, now we'll just use this image and
12:16we'll just send it over. So, this is the
12:18next segment of the b-roll and I'm using
12:20the previous image as a reference. Now,
12:22I'll generate the image of the pieces of
12:24b-roll that are in the script. We have
12:25them here. So, now I'll be creating the
12:27prompts to actually generate the
12:29footage. Previously, we generated the
12:30photos for each one of these segments
12:32and now we will be turning those into
12:34actual footage when the person is moving
12:36on screen and saying something. I'll be
12:38using Gemini Omni for generating this
12:40these segments. I want you to create me
12:43a detailed prompt for each one of these
12:46segments. I'll be uploading a photo that
12:48comes with each one of these as a
12:50reference so that everything stays
12:52consistent throughout the video.
12:54Generate me detailed prompt for each one
12:56of these segments so I can generate the
13:00full video out of it. All right, so
13:02look, this is very realistic image for
13:04the next segment. We'll continue with
13:06the next one. Again, it's very important
13:08to use the reference from the previous
13:09image to generate the next one. All
13:11right, so generated me the image here.
13:14Everything is good here besides the
13:15product because that's not the product
13:17that I want to use in the footage. This
13:19is why it's so important to use the
13:21right references. In this case, it
13:23didn't use the right product that is
13:25mine that I want to sell,
13:27and it just used the random product that
13:29it generated. Regenerate this image but
13:32with this product packaging instead of
13:35the bottle that you have there. So, now
13:36we have the right output. She has the
13:38product in hand. She has the soft gels
13:41in hand. She looks smiling. Everything
13:44is great. Very detailed high-quality
13:46image. Looks very realistic. And even if
13:49we zoom it in, it's still pretty
13:50realistic. Look at this. The detail here
13:54like on lips, on skin, on hair, even on
13:58the product. There is no misspellings in
14:00here. And here, look at the detail on
14:02her hands. Like even the lines. Like I
14:04can probably like read these lines and
14:06[ __ ] predict her fate. Now, in the
14:07last one, the reason I'm using this one
14:09as an inspiration instead of the one
14:12that is previous with the product in
14:13hand because it covers half of her face
14:16and I don't want it to generate some
14:18random, you know, face. Again, I want it
14:20to stay consistent throughout the video.
14:22So, I'll be using this one as a
14:24reference. Remember, we're generating
14:25prompts for video generation inside of
14:28the cloud and it [ __ ] up. It gave me
14:30like 9-second first one for the hook,
14:34but I only need 4 seconds. So, now I'll
14:36just clarify with cloud why that
14:38happened. So, now just take the script
14:40that we generated and I'll give it to
14:42cloud to generate me the prompts for
14:44each one of these segments that I can
14:46generate an actual footage with and then
14:49I'll upload the photos that we have
14:52generated inside of the charge GPT as a
14:54reference. So, it stays consistent
14:56throughout the whole video. When people
14:58say that you can generate a first video
15:00from the first try, they're full of
15:01[ __ ] Usually, it's iterations.
15:03Sometimes AI [ __ ] up. Sometimes you
15:06have to correct it. Sometimes you have
15:07to go back and recreate some of the
15:09elements. You have to recreate the
15:11prompts. You have to recreate the
15:12reference images. So, it's a back and
15:14forward work in order to get the perfect
15:16result. Now, it's important to do all of
15:18these steps manually at first so you
15:20understand the whole process and you
15:22understand the moving parts that come
15:23into creating a great video. But if you
15:25want to see how to automate this whole
15:26process inside of the Hermes agent, I'll
15:29record the whole process from start to
15:30finish how to automate this manual
15:33process into something that is
15:34automated, that is giving you a UGC
15:37videos every single day that are great
15:39quality. And if you want me to make that
15:41tutorial, like this video and comment
15:44agent below so I see there's enough
15:45demand. So now I uploaded this image
15:47with a lady inside of the car and I gave
15:49it the whole script from start to finish
15:52because I want my Claude to have the
15:53full context of where she is, what she
15:55does, how she looks like to create me a
15:58very detailed, very realistic prompts
16:00that I can use inside of the Google Omni
16:03to generate the actual footage that I'll
16:06use as my ads. In this particular
16:08segment, she has the sunscreen in hand,
16:12but in some other ones she would not
16:14have it. So I'll upload some of the
16:16other reference images so you see it as
16:18an example to generate me super detailed
16:21prompts that I can use inside of the key
16:23I E with Google Omni engine to generate
16:28the footage for each one of the
16:29segments. So another [ __ ] up here, the
16:31original script had a hook until the
16:33second four, but it turned out that it
16:36takes like nine seconds to say that hook
16:39out loud and that's why it had to
16:42generate this like longer prompt for the
16:44first segment of the video. So now we've
16:48generated the first part of the footage
16:50I inserted the prompt that Claude gave
16:52to me. It will generate the audio,
16:559 by 16 format because that's the best
16:59like vertical format that people look at
17:01on their phones.
17:02Duration in seconds. I mean, the script
17:05was about nine seconds long, so that's
17:07what we'll have to generate and we'll
17:09generate them in a MP4 format. I also
17:12uploaded the last frame here so that
17:14it's gradually moving into the second
17:17frame which creates that almost like
17:19it's impossible to see that it was
17:21created in segment. It looks like it was
17:23created in one go. And then I just click
17:25run. This generation alone cost me $4,
17:28which is a lot for such a small segment
17:30of just 9 seconds. So, let me see how it
17:32sounds like. It doesn't look so bad from
17:34the looking of it, but let's see.
17:37>> I'm a redhead.
17:38I burn in about 10 minutes.
17:41And it turns out I've been getting sun
17:42in my car every single day.
17:45>> It's not bad, right? Look look at the
17:46facial expressions here. I mean, it
17:48looks pretty like realistic. So, is it
17:51worth $4 just for this generation? It's
17:53pretty good. I mean, this this ad if if
17:56the script is actually good, I mean, it
17:57can [ __ ] print. So, now we're
17:59uploading the reference images for the
18:01first frame and last frame of the second
18:03segment and also the prompt here. We're
18:06keeping all of the settings the same, uh
18:09the resolution and the format here is
18:11adaptive because that's the only one
18:13that it will generate with. Now we're
18:15generating the second segment. Each one
18:17of these cost several dollars to
18:19generate. And so, when you combine all
18:21of it, let's say you're generating
18:22hundreds of videos per month, it comes
18:25down to thousands of dollars. And I know
18:27people that are spending 40, 50,000
18:29dollars a month just on HIXEL to
18:31generate these AI videos. However, if
18:34you think about the ROI from those, if
18:36you apply them properly, if you generate
18:37ads that are making you money, you're
18:40putting those ads in and you're getting
18:42money out, then it's actually a great
18:44ROI. And you could be spending that same
18:46money or even more on real photo shoots,
18:49on actors, on stuff that just like adds
18:51a lot of complexity to your business.
18:53So, we generated this in a draft format.
18:55It's 480 resolution. Let's see what it
18:57looks like.
19:00>> Your windshield blocks most UVA. Your
19:02side windows often block a lot less. And
19:05UVA is the sneaky one. No burn, no
19:08warning. It just goes deep and ages you.
19:11Google truck driver sun damage. 28 years
19:14behind the wheel and you can see exactly
19:16which side face the window.
19:17>> So now this is not perfect. The ending
19:19is definitely not what it should be, but
19:21this can be edited. Like very minimal
19:23editing. The rest of the script looks
19:25pretty nice and the quality is very
19:27good. The next part is done.
19:29>> So yes, sunscreen every day
19:32non-negotiable. But when you burn as
19:34fast as I do, you want back up.
19:37That's how I found astaxanthin. It's the
19:39red pigment and algae makes to survive
19:42brutal sunlight.
19:43That's literally why it's red.
19:45>> And one more segment that was generated
19:47here. And now she has a segment where
19:49the product is in the frame. So let's
19:51see what that looks like.
19:52>> There's a placebo-controlled trial out
19:54of Japan. People took it daily for about
19:579 weeks and their skin needed more UV
20:00before it turned red.
20:02Less moisture lost, too.
20:04>> All right. So now we have the footage
20:05with the product. Let's play it and see
20:07how it looks.
20:09>> This is Curcumin MD, 12 mg from the real
20:12algae, not synthetic, and it's an oil,
20:15which matters because it's fat soluble.
20:18>> Boom. All right. So now we have a great
20:20piece of footage here. This one doesn't
20:22look perfect. It has these like
20:24imperfections here, but like when you're
20:26looking on mobile, you don't really see
20:28that. But overall the quality here is
20:30pretty decent. Uh this was also
20:32generated in 480p format. It's not even
20:35like full format. So
20:37we would regenerate that in a proper
20:391080p,
20:41which is just highest quality possible
20:44and it will be looking much better. All
20:46right. So this is now the final product.
20:48My team made you some very slight edits.
20:50It took like 5 minutes and this is the
20:52ready product.
20:53>> I'm a redhead.
20:55I burn in about 10 minutes. And it turns
20:57out I've been getting sun in my car
20:59every single day.
21:00>> Okay. So now the the problem, right? So
21:03we're talking about the problem here.
21:04>> blocks most UVA. Your side windows often
21:07block a lot less.
21:08>> Emphasizing the problem, education.
21:10>> And UVA is the sneaky one. No burn, no
21:13warning. It just goes deep and ages you.
21:16>> Even more problems.
21:17>> Google truck driver sun damage.
21:19>> Even more problems.
21:20>> 28 years behind the
21:21>> Scared the [ __ ] out of them.
21:22>> wheel and you can see exactly which side
21:24face the window. So yes, sunscreen every
21:27day non-negotiable.
21:29>> Okay, so like typical options that
21:31people use for this. Oh, but I have a
21:32sunscreen. Yeah, but
21:34>> when you burn as fast as I do,
21:36you want backup. That's how I found
21:38astaxanthin.
21:39>> Boom.
21:40Introduction to the product 30 minutes
21:42at 30 seconds into the ad.
21:43>> It's the red pigment an algae makes to
21:45survive brutal sunlight.
21:48That's literally why it's red. This is
21:50>> Boom.
21:50>> Clarity MD 12 mil
21:52>> Introduction to the actual product.
21:53>> grams from the real algae, not
21:55synthetic. And it's
21:56>> Not synthetic, disqualifying other
21:58options.
21:58>> an oil, which matters because it's fat
22:00soluble.
22:01>> It doesn't replace sunscreen. It's
22:02support from the inside for every day
22:04you sat next to this window without
22:06thinking about it.
22:07>> Boom. And it's for everyone because
22:08everyone [ __ ] drives.
22:10>> Links
22:11below.
22:12>> Let's go. Very simple call to action.
22:13Great ad. Actually a bit more than 20
22:15minutes, but this was like a manual way
22:17just to show you guys how it works. In
22:19our business it's fully automated from
22:21start to finish. But here's a problem.
22:23Even if you have the best AI UGC in the
22:25world, right? And it's like it's
22:27amazing. But you send it to the funnel
22:29that does not convert, it wouldn't work.
22:31You'll be just burning money. Also, if
22:33you have this AI UGC like you know how
22:35to create it and you know how to do it
22:36yourself, but you don't have a team to
22:39help you execute it and come up with
22:41ideas and micro angles and having all of
22:43those processes solidified, then you'll
22:45be just forever stuck to your laptop.
22:47And people that have cracked this and
22:49scaled to $1 million a month, they
22:51operate by totally different rules. They
22:53don't just like manually create these
22:55videos themselves and trying to make it
22:57work. They have a process and a system
22:59for scaling their business and I
23:01outlined that whole process in a video
23:03that you can find in the description.
23:05This scaling blueprint will show you
23:06exactly how to scale your business to $1
23:08million a month and beyond. So, check it
23:10out below. Now, the best way to find an
23:11inspiration for creating these AI UGC
23:13videos because you don't want to just
23:15like create a random video. You want to
23:17take something that has been working in
23:19the market and recreate it for yourself,
23:21some proven winners from other people.
23:23And I have 283 of these bangers in one
23:26swipe file and I'll drop the link below
23:28this video as well, so you can check it
23:29out. So, for example, this ad has been
23:31running for 140 days. Like imagine how
23:35much money this one ad made. This is
23:37completely AI and look how simple it is.
23:39>> If you have one of these donut cushions,
23:41listen. They don't fix tailbone pain. I
23:44had three of them.
23:45They just shift the pressure to the ring
23:47around your tailbone.
23:48>> And remember, anytime you're trying to
23:50create a new ad, you need to have a
23:52specific micro angle in place in order
23:54for that ad to resonate with a specific
23:56consumer because otherwise, you'll be
23:58just burning your money. So, make sure
24:00to use the micro angles below to extract
24:02the micro angles for your product and
24:03I'll see you guys in the next video.

About this transcript

This page contains the full transcript of My AI UGC looks 100% real because I don't use Seedance (copy me) by Alex Fedotoff, generated from the public captions YouTube serves with the video. The transcript has 4,923 words across 695 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.