WEBVTT 00:00:00.000 --> 00:00:02.240 Think about the death of the Hollywood budget. 00:00:02.580 --> 00:00:06.759 It is a wild reality to consider. Picture this. 00:00:06.919 --> 00:00:09.919 Okay. You are waiting for the bus. You are holding 00:00:09.919 --> 00:00:12.519 absolutely nothing but your phone. Right. And 00:00:12.519 --> 00:00:14.980 you are producing a 30 -minute cinematic movie. 00:00:15.160 --> 00:00:17.800 Which sounds completely impossible. Right. It 00:00:17.800 --> 00:00:20.679 is in full 4K. The dialogue is perfectly lip 00:00:20.679 --> 00:00:23.839 -synced. Yeah. And your total budget is $0. It 00:00:23.839 --> 00:00:26.179 really does sound impossible. But it is happening 00:00:26.179 --> 00:00:27.800 right now. Yeah. I mean, people are actually 00:00:27.800 --> 00:00:30.960 bypassing the traditional studio system entirely. 00:00:32.079 --> 00:00:34.939 Welcome to today's Deep Dive. Thanks for having 00:00:34.939 --> 00:00:37.780 me. We are looking at something truly fascinating 00:00:37.780 --> 00:00:40.850 today. We really are. It is a mobile guide from 00:00:40.850 --> 00:00:44.750 March 2026. It's by Maxanne. Right. And it fundamentally 00:00:44.750 --> 00:00:47.250 changes everything about visual storytelling. 00:00:47.530 --> 00:00:49.789 It does. And what is really brilliant about it 00:00:49.789 --> 00:00:52.829 is the access. It completely removes the financial 00:00:52.829 --> 00:00:55.329 barrier to entry. Because historically, animation 00:00:55.329 --> 00:00:58.030 costs thousands of dollars per minute. Exactly. 00:00:58.030 --> 00:01:02.130 But now, the cost is just your time. Today, we 00:01:02.130 --> 00:01:05.420 are exploring a specific three -part... mobile 00:01:05.420 --> 00:01:08.400 pipeline. Yeah. The core stack. Right. First, 00:01:08.680 --> 00:01:11.760 we are going to architect the story in ChatGPT. 00:01:12.060 --> 00:01:15.219 Then we bring those static ideas to life with 00:01:15.219 --> 00:01:18.579 Grok AI. And then the final step. Right. Finally, 00:01:18.700 --> 00:01:21.280 we polish all the rough edges and cap cut. It 00:01:21.280 --> 00:01:23.980 is just three free tools on your phone. Yeah. 00:01:24.079 --> 00:01:27.420 But it represents one massive paradigm shift 00:01:27.420 --> 00:01:30.519 for creators. So let's unpack this step by step. 00:01:30.599 --> 00:01:32.340 Let's do it. We have to start with phase one. 00:01:33.439 --> 00:01:35.459 architecting the story right because you simply 00:01:35.459 --> 00:01:38.180 can't shoot a movie without a rock -solid script 00:01:38.180 --> 00:01:40.840 it is the essential first step absolutely we 00:01:40.840 --> 00:01:43.239 move from the grand promise of a mobile studio 00:01:43.239 --> 00:01:45.659 directly into the mind of the architect that 00:01:45.659 --> 00:01:48.239 is the perfect way to frame it before you even 00:01:48.239 --> 00:01:51.400 open a single video generation tool you need 00:01:51.400 --> 00:01:53.900 that underlying architecture because if you skip 00:01:53.900 --> 00:01:56.379 this you just get a folder of random clips exactly 00:01:56.829 --> 00:01:58.769 They look totally disconnected. The narrative 00:01:58.769 --> 00:02:01.129 just goes nowhere. The script gives the entire 00:02:01.129 --> 00:02:05.269 AI workflow its structural integrity. Yes, exactly. 00:02:05.530 --> 00:02:08.830 And in this pipeline, ChatGPT acts as your master 00:02:08.830 --> 00:02:11.530 showrunner. It wears a few different hats. It 00:02:11.530 --> 00:02:15.310 really does. It serves four very distinct vital 00:02:15.310 --> 00:02:18.509 roles. Okay. First, generating the overarching 00:02:18.509 --> 00:02:21.110 story. Right. Second, expanding the emotional 00:02:21.110 --> 00:02:24.560 details. Third. Writing the visual image prompts. 00:02:24.740 --> 00:02:27.479 And the fourth. Fourth is crafting the dialogue 00:02:27.479 --> 00:02:29.800 and the motion prompts. Let's pause on that very 00:02:29.800 --> 00:02:34.259 first step. Generating the story. Sure. The guide 00:02:34.259 --> 00:02:37.240 stresses asking for a multi -part story right 00:02:37.240 --> 00:02:40.020 up front. Yeah. That is a huge point. Why ask 00:02:40.020 --> 00:02:42.620 for multiple parts immediately? Because it establishes 00:02:42.620 --> 00:02:45.460 a massive, cohesive world right from the jump. 00:02:45.560 --> 00:02:47.680 Oh, I see. It builds a full content calendar 00:02:47.680 --> 00:02:50.319 from just a single chat session. That makes sense. 00:02:50.500 --> 00:02:53.780 You feed ChatGPT a clear concept up front. You 00:02:53.780 --> 00:02:56.120 tell it you want five distinct chapters. Right. 00:02:56.219 --> 00:02:58.699 And suddenly every single part is ready to become 00:02:58.699 --> 00:03:01.599 his own standalone video episode. That is serious 00:03:01.599 --> 00:03:03.960 efficiency. It really is. It keeps you from staring 00:03:03.960 --> 00:03:06.180 at a blank screen tomorrow. Right. It makes the 00:03:06.180 --> 00:03:09.599 heavy lifting worth it. But step two is where 00:03:09.599 --> 00:03:12.379 the actual cinematic magic happens. Expanding 00:03:12.379 --> 00:03:15.139 the emotion. Exactly. You have to force the AI 00:03:15.139 --> 00:03:18.219 to expand that core story. So you are asking 00:03:18.219 --> 00:03:21.639 ChatGPT to add more scenes? Yes. But more importantly, 00:03:21.699 --> 00:03:24.159 you are asking for deeper emotional moments. 00:03:24.219 --> 00:03:26.699 This is where beginners always fail. Really? 00:03:26.740 --> 00:03:30.520 Yeah. A thin, simple script leads directly to 00:03:30.520 --> 00:03:33.120 a thin, boring video. That makes total sense. 00:03:33.400 --> 00:03:35.259 When you spend a few extra prompts demanding 00:03:35.259 --> 00:03:38.340 emotional depth, You give the AI stylistic direction. 00:03:38.620 --> 00:03:41.659 It makes the final output feel moody and cinematic 00:03:41.659 --> 00:03:44.379 instead of flat and robotic. Right. The emotion 00:03:44.379 --> 00:03:47.580 dictates the pixels. Exactly. Then we move to 00:03:47.580 --> 00:03:50.639 step three, generating the image prompts. This 00:03:50.639 --> 00:03:53.120 is a crucial safety net for beginners. How so? 00:03:53.319 --> 00:03:56.780 You ask ChatGPT to basically act as your director 00:03:56.780 --> 00:03:59.300 of photography. Oh, wow. Yeah. It has to write 00:03:59.300 --> 00:04:01.659 a highly detailed visual description for every 00:04:01.659 --> 00:04:03.740 single scene. Like lighting and camera angles? 00:04:04.000 --> 00:04:07.050 Lighting, camera angles, weather. Everything. 00:04:07.150 --> 00:04:10.090 And then we reach step four. The big one. Yeah, 00:04:10.169 --> 00:04:13.069 this seems like the most critical piece. Generating 00:04:13.069 --> 00:04:15.490 the actual dialogue and motion prompts. This 00:04:15.490 --> 00:04:18.189 requires a very specific dual output from the 00:04:18.189 --> 00:04:21.009 AI for every scene. Dual output. Right. You need 00:04:21.009 --> 00:04:23.800 the motion prompt first. Okay. This explicitly 00:04:23.800 --> 00:04:26.379 explains how the camera and the environment should 00:04:26.379 --> 00:04:28.379 move. Give me an example of that. Something like 00:04:28.379 --> 00:04:31.800 camera slowly pushes in. While heavy dust particles 00:04:31.800 --> 00:04:34.740 drift across the darkened frame. You've got it. 00:04:34.819 --> 00:04:38.100 It adds that necessary atmosphere. Right. And 00:04:38.100 --> 00:04:41.240 the second required output is the dialogue script. 00:04:41.540 --> 00:04:44.000 What they actually say. Exactly. This is what 00:04:44.000 --> 00:04:46.740 each character actually says out loud. You keep 00:04:46.740 --> 00:04:49.300 this dual output format open on your screen. 00:04:49.980 --> 00:04:52.639 it becomes your working master script. It is 00:04:52.639 --> 00:04:55.220 like giving an architect not just the blueprints, 00:04:55.240 --> 00:04:57.879 but the exact brushstrokes for the painters. 00:04:58.100 --> 00:05:00.759 That is a great analogy. Be beat. Though I will 00:05:00.759 --> 00:05:02.540 admit something vulnerable here. What is that? 00:05:02.699 --> 00:05:05.279 I still wrestle with prompt drift myself when 00:05:05.279 --> 00:05:07.319 writing stories. Oh, it happens to absolutely 00:05:07.319 --> 00:05:10.040 everyone who uses these tools. It really does. 00:05:10.160 --> 00:05:12.139 For those who haven't experienced it, prompt 00:05:12.139 --> 00:05:16.310 drift is when AI... slowly forgets your initial 00:05:16.310 --> 00:05:19.310 instructions over time. The context window just 00:05:19.310 --> 00:05:21.730 gets overloaded. Yeah, it gets incredibly frustrating. 00:05:21.990 --> 00:05:24.670 It is a nightmare. So I have a probing question 00:05:24.670 --> 00:05:28.709 about that. Shoot. How do we prevent the AI from 00:05:28.709 --> 00:05:30.829 completely forgetting what the main character 00:05:30.829 --> 00:05:33.310 looks like between scene one and scene two? It 00:05:33.310 --> 00:05:35.889 all comes down to forcing visual anchors. Okay. 00:05:35.949 --> 00:05:39.810 Asking ChatGPT for strong, consistent visual 00:05:39.810 --> 00:05:42.709 character prompts for every single scene is absolutely 00:05:42.709 --> 00:05:45.490 mandatory. So you remind it every time? Every 00:05:45.490 --> 00:05:48.579 single time. You have to bake their physical 00:05:48.579 --> 00:05:51.439 description into every prong. If you don't, you 00:05:51.439 --> 00:05:53.540 end up with changing faces or totally different 00:05:53.540 --> 00:05:56.139 clothes. Which ruins the video. Completely. That 00:05:56.139 --> 00:05:59.399 breaks viewer immersion immediately. Right. Strong 00:05:59.399 --> 00:06:01.560 visual anchors keep the AI from changing their 00:06:01.560 --> 00:06:04.100 faces. Exactly. You can plug those precise descriptions 00:06:04.100 --> 00:06:08.079 into any image generator later. Or just use GPT's 00:06:08.079 --> 00:06:10.360 built -in image generator if you are on the free 00:06:10.360 --> 00:06:13.600 plan. So we have essentially built an incredibly 00:06:13.600 --> 00:06:16.560 detailed character locked script. We have the 00:06:16.560 --> 00:06:18.500 blueprint. But a text file doesn't entertain 00:06:18.500 --> 00:06:21.240 anyone. No, it doesn't. The hurdle now is visual 00:06:21.240 --> 00:06:24.019 execution. Right. Historically, that meant paying 00:06:24.019 --> 00:06:26.740 thousands of dollars for a rendering farm. Yeah, 00:06:26.779 --> 00:06:29.100 a rendering farm being massive computer networks 00:06:29.100 --> 00:06:31.920 used to process complex video. Right. We don't 00:06:31.920 --> 00:06:34.000 have access to servers like that. We just have 00:06:34.000 --> 00:06:37.379 a phone. And this is where the massive March 00:06:37.379 --> 00:06:41.920 2026 update to the Grok app. completely changes 00:06:41.920 --> 00:06:44.800 the game. Tell me about that. Specifically, the 00:06:44.800 --> 00:06:47.560 animate photo feature. This is the core trick 00:06:47.560 --> 00:06:50.540 of the entire pipeline. Okay. It takes your static 00:06:50.540 --> 00:06:53.639 mid -journey or GPT image and turns it into a 00:06:53.639 --> 00:06:57.000 fluid six -second video. And it has fully integrated 00:06:57.000 --> 00:07:00.389 lip sync built right in. Yeah. It is wild. You 00:07:00.389 --> 00:07:02.970 access it by simply opening the Grok app on your 00:07:02.970 --> 00:07:05.550 phone. You tap the Imagine tab at the bottom. 00:07:05.649 --> 00:07:08.209 Right. Then you tap the little photo icon and 00:07:08.209 --> 00:07:10.430 choose animate photo. And just a quick note for 00:07:10.430 --> 00:07:12.129 you listening, if you don't see that button, 00:07:12.230 --> 00:07:14.670 go update or reinstall the app. You need that 00:07:14.670 --> 00:07:18.550 specific March 2026 version. Right. Once you 00:07:18.550 --> 00:07:21.129 are in the interface, you import your static 00:07:21.129 --> 00:07:24.009 scene image. Okay. Then you write your prompt. 00:07:24.329 --> 00:07:27.110 Using the master script. Exactly. You're combining 00:07:27.110 --> 00:07:30.850 that motion prompt ChatGPT gave you earlier with 00:07:30.850 --> 00:07:33.189 your character's dialogue. You just type the 00:07:33.189 --> 00:07:36.910 exact spoken dialogue in quotation marks. So 00:07:36.910 --> 00:07:39.129 the AI reads the text and maps it to the mouth. 00:07:39.329 --> 00:07:41.410 Yeah, it does. But there are some pretty severe 00:07:41.410 --> 00:07:43.889 constraints here, right? Huge constraints. You 00:07:43.889 --> 00:07:46.850 must limit the generation to two dialogue exchanges 00:07:46.850 --> 00:07:49.990 per prompt. Two exchanges. That is the absolute 00:07:49.990 --> 00:07:52.629 maximum. What happens if you do more? if you 00:07:52.629 --> 00:07:55.329 feed it a massive block of text the processing 00:07:55.329 --> 00:07:58.670 load gets too heavy okay the ai loses coherence 00:07:58.670 --> 00:08:01.449 and characters just skip lines completely so 00:08:01.449 --> 00:08:04.629 how do we actually get past six seconds six seconds 00:08:04.629 --> 00:08:07.389 is a neat trick but it is not a television episode 00:08:07.389 --> 00:08:11.170 that is the ultimate secret sauce of 2026 which 00:08:11.170 --> 00:08:13.569 is the extend feature okay after you generate 00:08:13.569 --> 00:08:16.930 that first flawless six second clip you tap the 00:08:16.930 --> 00:08:19.870 three dots in the corner you hit extend then 00:08:19.870 --> 00:08:23.920 you paste The next small chunk of your dialogue. 00:08:24.160 --> 00:08:26.379 And you just generate it again. Right. The clip 00:08:26.379 --> 00:08:29.439 seamlessly grows from six seconds to 12 seconds. 00:08:29.620 --> 00:08:32.139 Wow. It takes the last frame of the first clip 00:08:32.139 --> 00:08:34.460 and uses it as the starting point for the next. 00:08:34.539 --> 00:08:36.299 So you just keep doing that. You repeat that 00:08:36.299 --> 00:08:39.159 process. It goes to 18 seconds. Yeah. Then 24. 00:08:39.659 --> 00:08:43.360 Then 30 seconds. You stitch them together block 00:08:43.360 --> 00:08:46.059 by block. It is like stacking Lego blocks of 00:08:46.059 --> 00:08:51.120 data. Exactly. To sex islands. Whoa. Imagine 00:08:51.120 --> 00:08:53.779 scaling a single photo into a whole half hour 00:08:53.779 --> 00:08:56.840 animated film. It is a massive, unprecedented 00:08:56.840 --> 00:08:59.809 shift in solo production. Yeah. But there is 00:08:59.809 --> 00:09:02.330 a highly critical insight you must follow to 00:09:02.330 --> 00:09:04.870 make it work. What is that? It is the defining 00:09:04.870 --> 00:09:08.669 differentiator for creators in 2026. The guide 00:09:08.669 --> 00:09:11.070 calls it character parenting. Character parenting. 00:09:11.389 --> 00:09:13.269 Yeah. What does that actually mean in practice? 00:09:13.629 --> 00:09:16.389 In your Grok generation prompts, you always put 00:09:16.389 --> 00:09:18.730 the character's gender and rough age in parentheses 00:09:18.730 --> 00:09:21.929 right next to their name. For example, you literally 00:09:21.929 --> 00:09:27.049 type Amina, open parenthesis, woman, comma. Why 00:09:27.049 --> 00:09:31.149 is that tiny detail so profoundly important? 00:09:31.549 --> 00:09:33.789 Because we have to understand how these AI models 00:09:33.789 --> 00:09:36.789 are trained. Right. The latent space is heavily 00:09:36.789 --> 00:09:40.419 biased toward Western media. Okay. Latent space 00:09:40.419 --> 00:09:43.519 bias being AI models favoring certain types of 00:09:43.519 --> 00:09:46.240 training data. Exactly. If you just type a non 00:09:46.240 --> 00:09:48.740 -Western name like Amina, the AI searches its 00:09:48.740 --> 00:09:51.220 database and gets confused. Oh, I see. It leads 00:09:51.220 --> 00:09:54.860 to default generic voices. Or worse, it creates 00:09:54.860 --> 00:09:57.519 entirely wrong facial expressions that erase 00:09:57.519 --> 00:10:00.059 ethnic features. It basically loses its anchor 00:10:00.059 --> 00:10:01.980 and starts flipping voices during the extension 00:10:01.980 --> 00:10:04.919 process. Exactly. Adding that simple demographic 00:10:04.919 --> 00:10:06.980 label acts as an anchor weight in the neural 00:10:06.980 --> 00:10:09.720 network. It stops the AI from... drifting or 00:10:09.720 --> 00:10:11.879 messing up the cultural nuances. It saves time. 00:10:12.019 --> 00:10:14.460 Saves you hours of frustrating regeneration time. 00:10:14.620 --> 00:10:17.100 Let me push back here a bit. Sure. The tech sounds 00:10:17.100 --> 00:10:19.659 amazing, but if I am stitching together these 00:10:19.659 --> 00:10:22.480 small six -second clips, isn't the audio going 00:10:22.480 --> 00:10:24.539 to sound incredibly disjointed at the seams? 00:10:24.799 --> 00:10:27.460 That is a fair concern. Plus, what happens when 00:10:27.460 --> 00:10:30.580 the clip inevitably goes off script or hallucinates 00:10:30.580 --> 00:10:33.399 and the mouth just stops sinking? Those are the 00:10:33.399 --> 00:10:35.899 exact right questions. Hallucinations being... 00:10:36.120 --> 00:10:39.080 When the AI invents random incorrect details 00:10:39.080 --> 00:10:41.720 that break the scene. Right. It definitely happens. 00:10:41.899 --> 00:10:44.639 Yeah. The audio usually smooths out because grok 00:10:44.639 --> 00:10:47.440 overlaps the audio frame slightly. That's clever. 00:10:47.679 --> 00:10:51.320 But when the visual completely breaks, you shouldn't 00:10:51.320 --> 00:10:53.960 just spam the regenerate button. Why not? That 00:10:53.960 --> 00:10:57.019 wastes your daily limits. First, you need to 00:10:57.019 --> 00:10:59.330 check the fundamental basics. Like what? Check 00:10:59.330 --> 00:11:01.129 that your character labels are actually present 00:11:01.129 --> 00:11:03.330 in the prompt. Make sure you didn't accidentally 00:11:03.330 --> 00:11:06.149 delete the parentheses. Right. That is usually 00:11:06.149 --> 00:11:08.429 the culprit. Okay. And second, make sure you 00:11:08.429 --> 00:11:10.429 haven't overloaded the context window. Check 00:11:10.429 --> 00:11:13.210 the dialogue length. Check if your dialogue snippet 00:11:13.210 --> 00:11:17.289 is too long. Over 90 % of bad, hallucinated clips 00:11:17.289 --> 00:11:20.789 come from those two specific user errors. Got 00:11:20.789 --> 00:11:23.730 it. Fix labels. Cut dialogue short, then try 00:11:23.730 --> 00:11:26.250 generating it again. That is the exact troubleshooting 00:11:26.250 --> 00:11:28.629 loop. Okay. Once the clip generates cleanly, 00:11:28.629 --> 00:11:30.629 you save it to your camera roll. Then you move 00:11:30.629 --> 00:11:32.830 to the very next scene in your script. And just 00:11:32.830 --> 00:11:35.230 repeat. You keep systematically repeating this 00:11:35.230 --> 00:11:37.750 until the full story is done. And the guide mentions 00:11:37.750 --> 00:11:40.570 a massive time -saving trick here. Yeah, it is 00:11:40.570 --> 00:11:42.470 brilliant. What is it? If two back -to -back 00:11:42.470 --> 00:11:45.049 scenes use the exact same setting and the same 00:11:45.049 --> 00:11:47.870 characters, just reuse the same starting image. 00:11:48.090 --> 00:11:51.330 Oh, right. Don't prompt a brand new one. It saves 00:11:51.330 --> 00:11:53.990 rendering time and it keeps your visual style 00:11:53.990 --> 00:11:56.690 highly consistent across the episode. OK, let's 00:11:56.690 --> 00:11:59.090 take a breath. Yeah. We now have a camera roll 00:11:59.090 --> 00:12:03.230 completely full of extended 30 second video clips. 00:12:03.590 --> 00:12:07.529 But a crowded folder of clips isn't a movie. 00:12:08.129 --> 00:12:11.850 We need to cut the robotic fat. And that brings 00:12:11.850 --> 00:12:14.429 us to the final crucial step. We are going to 00:12:14.429 --> 00:12:16.149 take a quick break. Right. Stick around. And 00:12:16.149 --> 00:12:19.289 we are back. So before the break. We generated 00:12:19.289 --> 00:12:22.850 all our raw footage. We did. Now we are moving 00:12:22.850 --> 00:12:25.870 all those assets into CapCut. It is a totally 00:12:25.870 --> 00:12:28.690 free mobile video editor. Right. You start a 00:12:28.690 --> 00:12:31.070 brand new project on your phone. Okay. You import 00:12:31.070 --> 00:12:33.889 all your saved clips from the camera roll. Because 00:12:33.889 --> 00:12:36.909 you generated them in order, they usually drop 00:12:36.909 --> 00:12:39.210 into the timeline sequentially. I want to clarify 00:12:39.210 --> 00:12:41.409 something for you listening. Yeah. The focus 00:12:41.409 --> 00:12:44.860 in this stage. isn't fancy transitions. We aren't 00:12:44.860 --> 00:12:48.440 doing crazy MTV -style edits here. No, the entire 00:12:48.440 --> 00:12:51.799 philosophy of this edit is purely about hiding 00:12:51.799 --> 00:12:54.960 the AI's technical flaws. Okay. You have to scrub 00:12:54.960 --> 00:12:58.100 through each clip very carefully. You are hunting 00:12:58.100 --> 00:13:00.879 for two specific immersion -breaking issues. 00:13:01.220 --> 00:13:03.139 What exactly are we hunting for on that timeline? 00:13:03.480 --> 00:13:05.779 First... dead spots where the character just 00:13:05.779 --> 00:13:08.860 freezes mid -scene. This usually happens right 00:13:08.860 --> 00:13:10.639 around the mouth when the line of dialogue finishes 00:13:10.639 --> 00:13:13.879 early. Second, you are looking for moments where 00:13:13.879 --> 00:13:16.480 the video keeps running awkwardly after the intended 00:13:16.480 --> 00:13:19.379 emotion ends. So you have to meticulously pinch 00:13:19.379 --> 00:13:22.000 and zoom on that timeline to trim those clips. 00:13:22.259 --> 00:13:25.159 Exactly. You use CapCut's basic split and trim 00:13:25.159 --> 00:13:28.700 tools. You ruthlessly cut out those empty robotic 00:13:28.700 --> 00:13:31.879 moments. You add a simple text title at the beginning 00:13:31.879 --> 00:13:34.620 of the project. And once the flow feels naturally 00:13:34.620 --> 00:13:38.259 human, you export the whole thing in 1080p. Let's 00:13:38.259 --> 00:13:40.740 do a harsh reality check on the actual time and 00:13:40.740 --> 00:13:43.320 output here. Okay. How long does this entirely 00:13:43.320 --> 00:13:46.240 mobile process actually take? Let's break down 00:13:46.240 --> 00:13:49.039 the basic math. Sure. If you have a script with 00:13:49.039 --> 00:13:52.759 10 scenes and you do 6 to 10 extensions for each 00:13:52.759 --> 00:13:54.580 of those scenes. That is a lot of extension. 00:13:54.840 --> 00:13:59.120 It is. But you end up with a solid. 25 to 35 00:13:59.120 --> 00:14:02.559 minute video. That is literally a full television 00:14:02.559 --> 00:14:06.759 episode length. It is. And the total actual production 00:14:06.759 --> 00:14:09.159 time for that. Yeah. It is sitting right around 00:14:09.159 --> 00:14:12.179 two to three hours. Wow. That is from writing 00:14:12.179 --> 00:14:15.299 the initial chat GPT prompt all the way to the 00:14:15.299 --> 00:14:18.019 final exported video. I have to say two to three 00:14:18.019 --> 00:14:21.000 hours sounds incredibly fast for 30 minutes of 00:14:21.000 --> 00:14:24.620 pristine animation. It does. But I want to point 00:14:24.620 --> 00:14:27.299 out to you listening. Yeah. This requires serious, 00:14:27.340 --> 00:14:30.279 hyper -focused work. It isn't just a push a button 00:14:30.279 --> 00:14:32.360 and take a nap kind of magic trick. No, it is 00:14:32.360 --> 00:14:34.860 not. If you are thinking trimming microseconds 00:14:34.860 --> 00:14:37.740 of dead air on a tiny phone screen sounds like 00:14:37.740 --> 00:14:39.960 a nightmare, you are not entirely wrong. Right. 00:14:40.039 --> 00:14:43.419 It is meticulous. It is highly meticulous. You 00:14:43.419 --> 00:14:46.919 are actively directing the AI. You are extending 00:14:46.919 --> 00:14:50.059 clips block by block. You are constantly troubleshooting 00:14:50.059 --> 00:14:52.759 broken prompts. You are trimming single frames. 00:14:52.879 --> 00:14:55.759 It is real demanding creative work. It is just 00:14:55.759 --> 00:14:57.639 a totally different kind of work than traditional 00:14:57.639 --> 00:15:00.960 animating. Exactly. But I have to ask, are those 00:15:00.960 --> 00:15:04.200 tiny dead spots or frozen mouths really that 00:15:04.200 --> 00:15:07.139 big of a deal for the final viewer? Yes. Can't 00:15:07.139 --> 00:15:09.379 we just leave them in to save an hour of editing? 00:15:09.759 --> 00:15:11.740 They are a massive deal. We are dealing with 00:15:11.740 --> 00:15:14.179 the uncanny valley here. Right. Leaving even 00:15:14.179 --> 00:15:18.240 a single second of a frozen, lifeless AI face. 00:15:18.940 --> 00:15:20.860 Instantly shatters the viewer's psychological 00:15:20.860 --> 00:15:23.840 immersion. It just feels wrong. It deeply unsettles 00:15:23.840 --> 00:15:26.059 the human brain. It will completely ruin your 00:15:26.059 --> 00:15:28.759 audience retention. People will click away immediately. 00:15:29.139 --> 00:15:31.399 Yeah, dead air shatters the illusion. Trimming 00:15:31.399 --> 00:15:34.360 hides the robotic awkwardness completely. That 00:15:34.360 --> 00:15:37.980 meticulousness is the exact barrier that separates 00:15:37.980 --> 00:15:40.860 professional -looking storytelling from the vast 00:15:40.860 --> 00:15:46.740 sea of lazy, low -effort AI spam online. Trimming 00:15:46.740 --> 00:15:49.700 isn't optional. It is essential. So looking at 00:15:49.700 --> 00:15:52.980 the big picture, who is this mobile pipeline 00:15:52.980 --> 00:15:56.720 actually built for? It is heavily optimized for 00:15:56.720 --> 00:16:00.620 a few specific types of creators. OK. It is absolutely 00:16:00.620 --> 00:16:03.740 perfect for faceless YouTube creators. Right. 00:16:03.820 --> 00:16:06.320 People who want to publish long form stories 00:16:06.320 --> 00:16:09.480 consistently without ever putting their real 00:16:09.480 --> 00:16:12.620 face on camera. The guide also explicitly mentions 00:16:12.620 --> 00:16:16.000 cultural storytellers. Yes. This is profound. 00:16:16.559 --> 00:16:20.639 If you want to produce localized drama, rich 00:16:20.639 --> 00:16:23.320 historical folklore, or deeply serialized fiction, 00:16:23.580 --> 00:16:25.940 this workflow is exactly what you've been waiting 00:16:25.940 --> 00:16:28.860 for. And it is built entirely for beginners with 00:16:28.860 --> 00:16:31.059 absolutely zero budget. Because every single 00:16:31.059 --> 00:16:33.679 tool in this impressive stack is entirely free. 00:16:33.840 --> 00:16:36.740 Right. And remember that multi -part story format 00:16:36.740 --> 00:16:39.419 we forced ChatGPT to write in step one? Yeah, 00:16:39.419 --> 00:16:41.320 the calendar. That builds natural subscriber 00:16:41.320 --> 00:16:43.539 retention. Each episode ends with a scripted 00:16:43.539 --> 00:16:45.519 cliffhanger. Oh, that's smart. The next part 00:16:45.519 --> 00:16:47.139 is already written on your phone. The audience 00:16:47.139 --> 00:16:50.019 has a compelling reason to return tomorrow. Let's 00:16:50.019 --> 00:16:52.259 synthesize the core takeaway from all of this. 00:16:52.399 --> 00:16:56.220 Two sec silence. The historical financial barrier 00:16:56.220 --> 00:16:59.399 to cinematic storytelling has effectively vanished. 00:16:59.679 --> 00:17:01.940 It really has evaporated overnight. By intentionally 00:17:01.940 --> 00:17:05.740 stacking ChatGPT, Grok's new update, and CapCut 00:17:05.740 --> 00:17:09.160 together, an entire world -class animation studio 00:17:09.160 --> 00:17:12.849 now fits quietly in your pocket. It is amazing. 00:17:13.130 --> 00:17:15.710 For zero dollars, you just build the world scene 00:17:15.710 --> 00:17:18.650 by scene, extension by extension. It is a profound 00:17:18.650 --> 00:17:21.309 tectonic shift in media. Yeah. The technical 00:17:21.309 --> 00:17:23.750 tools are no longer the obstacle keeping you 00:17:23.750 --> 00:17:26.349 from creating. The only thing you actually need 00:17:26.349 --> 00:17:29.029 to bring to the table now is a story that is 00:17:29.029 --> 00:17:31.230 actually worth telling. That is a really beautiful 00:17:31.230 --> 00:17:34.130 way to frame it. It makes you wonder. About what? 00:17:34.230 --> 00:17:36.690 If the financial cost of rendering breathtaking 00:17:36.690 --> 00:17:39.990 animation is now completely zero, will the value 00:17:39.990 --> 00:17:43.289 of deeply human, unique cultural folklore actually 00:17:43.289 --> 00:17:46.450 skyrocket? I believe so. Right. When the execution 00:17:46.450 --> 00:17:49.650 becomes free and effortless, genuine human originality 00:17:49.650 --> 00:17:53.430 becomes absolutely priceless. Exactly. So what 00:17:53.430 --> 00:17:56.130 deeply personal story have you been holding back 00:17:56.130 --> 00:17:58.069 because you thought you couldn't afford to tell 00:17:58.069 --> 00:18:01.569 it? Two sec silence. That is the real question. 00:18:01.789 --> 00:18:03.569 The next time you are just waiting for the bus, 00:18:03.730 --> 00:18:05.809 remember. You have a massive Hollywood studio 00:18:05.809 --> 00:18:07.950 sitting right there in your pocket. Thank you 00:18:07.950 --> 00:18:10.809 for exploring this wild new frontier with us 00:18:10.809 --> 00:18:12.890 today. Thanks for having me. Thanks for joining 00:18:12.890 --> 00:18:14.849 the Deep Dive. We will see you next time.