WEBVTT 00:00:00.000 --> 00:00:02.459 Imagine for a moment the cost of producing a 00:00:02.459 --> 00:00:04.700 high -quality professional 30 -second video. 00:00:06.139 --> 00:00:07.860 What do you think that would typically set you 00:00:07.860 --> 00:00:10.160 back? Well, with what we're about to dig into, 00:00:10.359 --> 00:00:13.279 maybe just 13 cents. 13 cents, wow. You know, 00:00:13.279 --> 00:00:14.740 when I first heard that, I almost didn't believe 00:00:14.740 --> 00:00:17.640 it. That really does feel like a huge shift, 00:00:17.699 --> 00:00:19.719 like a seismic shift underway in how we create 00:00:19.719 --> 00:00:22.699 video. Welcome to the Deep Dive. Yeah, today 00:00:22.699 --> 00:00:25.679 we're plunging into this remarkable world of 00:00:25.679 --> 00:00:28.039 automated AI video production. We're talking... 00:00:28.570 --> 00:00:31.429 unbelievably low costs, totally transforming 00:00:31.429 --> 00:00:33.770 how we think about content. So our mission on 00:00:33.770 --> 00:00:36.130 this deep dive really is to unpack how a new 00:00:36.130 --> 00:00:40.090 AI model, it's called C -Dance, when you combine 00:00:40.090 --> 00:00:42.890 it with smart automation platforms like NEN. 00:00:43.490 --> 00:00:45.929 How it's making professional looking video content, 00:00:46.049 --> 00:00:47.950 well, practically free for pretty much everyone. 00:00:48.250 --> 00:00:50.170 Exactly. We'll kind of compare the old costs 00:00:50.170 --> 00:00:52.689 to this new reality, lay out the blueprint for, 00:00:52.750 --> 00:00:55.390 you know, building your own AI studio, explore 00:00:55.390 --> 00:00:58.289 each department, if you will, of this automated 00:00:58.289 --> 00:01:01.170 setup. And then, yeah, crucially, look at the 00:01:01.170 --> 00:01:02.890 final budget and what this all means, the big 00:01:02.890 --> 00:01:04.849 implications for creators and businesses everywhere. 00:01:05.250 --> 00:01:09.250 OK, so let's set the scene a bit. Think back 00:01:09.250 --> 00:01:12.189 just a few years, maybe even just like months 00:01:12.189 --> 00:01:15.140 ago. If you wanted a high -quality video, you 00:01:15.140 --> 00:01:16.340 weren't just grabbing your phone, right? Oh, 00:01:16.340 --> 00:01:18.680 no way. You had to invest in a good camera, often 00:01:18.680 --> 00:01:21.079 thousands of dollars. Then came the lights, the 00:01:21.079 --> 00:01:23.879 microphones. You'd spend hours shooting, probably, 00:01:24.060 --> 00:01:26.319 followed by days hunched over some complex editing 00:01:26.319 --> 00:01:29.579 software, painstakingly piecing clips together, 00:01:29.700 --> 00:01:32.359 adding sound, fiddling with colors. The time, 00:01:32.359 --> 00:01:35.150 the money, the expertise needed. It was, well, 00:01:35.189 --> 00:01:37.090 significant. That really was the old world. Yeah. 00:01:37.290 --> 00:01:39.849 Then, you know, the first wave of AI videos started 00:01:39.849 --> 00:01:42.409 arriving. Companies like Google, they unveiled 00:01:42.409 --> 00:01:45.150 these powerful models. Take VO3, for instance. 00:01:45.329 --> 00:01:47.170 They were kind of like these exclusive Hollywood 00:01:47.170 --> 00:01:49.670 studios. They could produce stunning, really 00:01:49.670 --> 00:01:51.689 cinematic clips that were like impossible before 00:01:51.689 --> 00:01:54.510 without massive budgets and crews. But much like 00:01:54.510 --> 00:01:56.989 those big studios, the cost kept them out of 00:01:56.989 --> 00:01:58.849 reach for most people. It was really a tool for 00:01:58.849 --> 00:02:02.480 big corporations. Right. But now. It feels like 00:02:02.480 --> 00:02:04.359 the entire landscape has just been turned upside 00:02:04.359 --> 00:02:05.840 down. And here's where it gets really interesting, 00:02:05.939 --> 00:02:09.300 I think. A new player, this AI video model called 00:02:09.300 --> 00:02:12.819 Seedence, has entered the scene. Think of it 00:02:12.819 --> 00:02:15.419 maybe as a brilliant, really efficient independent 00:02:15.419 --> 00:02:17.740 film studio that just opened up down the street. 00:02:18.539 --> 00:02:22.639 And according to the most respected industry 00:02:22.639 --> 00:02:26.180 benchmarks, its films are actually better than 00:02:26.180 --> 00:02:28.039 the ones coming out of those big Hollywood AI 00:02:28.039 --> 00:02:30.710 studios. And here's the bombshell, yeah. The 00:02:30.710 --> 00:02:33.270 seismic shift that changes absolutely everything. 00:02:33.449 --> 00:02:36.030 This new studio. It's not just better. It's 25 00:02:36.030 --> 00:02:38.030 times cheaper. This isn't just, you know, an 00:02:38.030 --> 00:02:40.069 incremental update. This is a complete revolution. 00:02:40.289 --> 00:02:42.729 It's like going from hand -cranked engines straight 00:02:42.729 --> 00:02:45.009 to electric cars overnight. That's incredible. 00:02:45.569 --> 00:02:48.050 So when you look back at that old landscape of 00:02:48.050 --> 00:02:50.539 professional video. What's the single biggest 00:02:50.539 --> 00:02:53.280 barrier you feel this new model just absolutely 00:02:53.280 --> 00:02:56.060 shattered like overnight? Oh, without a doubt, 00:02:56.120 --> 00:02:59.099 it's the monumental barrier of cost for decades. 00:02:59.319 --> 00:03:01.560 Right. High quality video was just inaccessible 00:03:01.560 --> 00:03:03.319 for most people. It was a luxury for those with 00:03:03.319 --> 00:03:06.620 huge budgets. That gate has just been blown wide 00:03:06.620 --> 00:03:09.719 open. So. Connecting this to the biggle picture, 00:03:10.020 --> 00:03:12.860 how do we really know seed ends is better than, 00:03:12.879 --> 00:03:16.099 say, a powerhouse like Google's VO3? Well, the 00:03:16.099 --> 00:03:17.939 answer comes from the system called the Arena 00:03:17.939 --> 00:03:20.900 ELO rankings. Ah, right. Like think of it maybe 00:03:20.900 --> 00:03:23.280 like the ranking system in professional chess, 00:03:23.379 --> 00:03:25.819 but for AI models instead. Exactly. They show 00:03:25.819 --> 00:03:28.780 two AI models to a human voter who picks which 00:03:28.780 --> 00:03:31.639 one did a better job. based on a prompt, but 00:03:31.639 --> 00:03:33.460 they don't know which model made which video. 00:03:33.659 --> 00:03:36.479 So it's a proper blind taste test. Over 20 ,000 00:03:36.479 --> 00:03:38.460 matchups, I think. Yeah, a huge number. And after 00:03:38.460 --> 00:03:40.439 tens of thousands of these head -to -head matchups, 00:03:40.659 --> 00:03:42.860 Seedance has officially, you know, dethroned 00:03:42.860 --> 00:03:46.360 VO3. People consistently and crucially, without 00:03:46.360 --> 00:03:49.840 bias, prefer Seedance's video quality, the coherence, 00:03:49.879 --> 00:03:52.560 the artistic style. And what's fascinating here, 00:03:52.639 --> 00:03:54.039 right, is that a new champion that's slightly 00:03:54.039 --> 00:03:55.520 better, that's just an interesting development, 00:03:55.740 --> 00:03:58.000 but a champion that's better and exponentially 00:03:58.000 --> 00:04:00.710 cheaper, that's the revolution. Okay, so let's 00:04:00.710 --> 00:04:02.370 talk numbers then, because they really are the 00:04:02.370 --> 00:04:04.370 heart of this story, aren't they? When you actually 00:04:04.370 --> 00:04:06.530 integrate these models into an automation workflow, 00:04:06.870 --> 00:04:10.250 the costs are just staggering in their difference. 00:04:10.449 --> 00:04:12.990 They really are. So Google's VO3, it charges 00:04:12.990 --> 00:04:16.110 a pretty staggering $0 .75 per second of generated 00:04:16.110 --> 00:04:20.750 video. CDance at 720p resolution is just $0 .06 00:04:20.750 --> 00:04:23.649 per second. That's over 12 times cheaper already. 00:04:24.490 --> 00:04:27.449 But it gets even wilder. At 480p resolution, 00:04:27.910 --> 00:04:31.230 CDance is an almost unbelievable $0 .003 per 00:04:31.230 --> 00:04:34.350 second. Wow. Yeah. That's 25 times cheaper than 00:04:34.350 --> 00:04:36.050 the previous market leap. Okay, let's put that 00:04:36.050 --> 00:04:38.550 in real world terms for you listening. Imagine 00:04:38.550 --> 00:04:40.790 you want to create a 30 -second video clip, maybe 00:04:40.790 --> 00:04:43.170 for social media, TikTok, whatever. With VO3, 00:04:43.290 --> 00:04:46.449 that single clip would cost you $22 .50. With 00:04:46.449 --> 00:04:48.790 CDance, the exact same 30 -second clip costs 00:04:48.790 --> 00:04:52.019 you just $0 .90. 96. I mean, for the price of 00:04:52.019 --> 00:04:54.579 one video from that old, expensive studio, you 00:04:54.579 --> 00:04:57.300 can now produce 25 films with your new independent 00:04:57.300 --> 00:04:59.959 setup. Whoa. Yeah, imagine being able to A -B 00:04:59.959 --> 00:05:02.860 test dozens of video ad concepts, right? Or scaling 00:05:02.860 --> 00:05:04.879 your content creation to target these, like, 00:05:04.920 --> 00:05:07.800 ultra -niche audiences all for pennies. We're 00:05:07.800 --> 00:05:10.259 already seeing things like new AI -powered ASMR 00:05:10.259 --> 00:05:13.639 channels pop up, reaching millions of views with 00:05:13.639 --> 00:05:16.459 just a handful of videos built entirely with 00:05:16.459 --> 00:05:18.819 these kinds of automated systems. Yeah. The barrier 00:05:18.819 --> 00:05:21.300 to entry for truly ambitious content is just, 00:05:21.319 --> 00:05:24.160 well, it's gone. So beyond just saving money, 00:05:24.319 --> 00:05:26.959 which is obviously huge, what's the actual sort 00:05:26.959 --> 00:05:29.879 of seismic shift this enables for, say, a solo 00:05:29.879 --> 00:05:32.790 creator? Or maybe even a small team. It fundamentally 00:05:32.790 --> 00:05:35.850 unlocks strategies previously just reserved for 00:05:35.850 --> 00:05:38.930 big corporations. OK, now that we kind of grasp 00:05:38.930 --> 00:05:41.850 the sheer scale of this change, let's talk practicalities. 00:05:41.990 --> 00:05:44.449 How do you actually build this studio? You know, 00:05:44.449 --> 00:05:46.610 how do you harness this incredible power? Right. 00:05:46.689 --> 00:05:49.129 So you build your own automated film studio using 00:05:49.129 --> 00:05:52.060 a platform like NAN. Think of an NEN workflow 00:05:52.060 --> 00:05:55.920 as like a digital assembly line. It connects 00:05:55.920 --> 00:05:57.899 different tools and services so they work together 00:05:57.899 --> 00:05:59.899 automatically without needing to write any code, 00:05:59.980 --> 00:06:02.819 which is key for most people. Our AI film studio 00:06:02.819 --> 00:06:04.620 workflow, you can basically break it down into 00:06:04.620 --> 00:06:06.959 four distinct departments, you could say. Each 00:06:06.959 --> 00:06:08.680 one's responsible for a different part of the 00:06:08.680 --> 00:06:11.060 production process. Okay, so the first one is 00:06:11.060 --> 00:06:14.019 the writer's room. This handled the initial creative 00:06:14.019 --> 00:06:16.560 spark, the input. This is where your video ideas 00:06:16.560 --> 00:06:18.439 are born and kind of shaped up. Then you've got 00:06:18.439 --> 00:06:21.120 the sound stage. This handles the clip generation. 00:06:21.660 --> 00:06:24.699 Here, the AI director takes those ideas and actually 00:06:24.699 --> 00:06:27.439 shoots the video scenes. Following that is the 00:06:27.439 --> 00:06:29.180 sound department for sound generation, where 00:06:29.180 --> 00:06:32.040 a specialized audio AI adds all the effects and 00:06:32.040 --> 00:06:34.680 music. Right. And then finally, the editing bay. 00:06:35.069 --> 00:06:37.290 This is where a master editor AI takes all the 00:06:37.290 --> 00:06:39.930 film scenes and the audio and just assembles 00:06:39.930 --> 00:06:42.709 them into a single polished final cut. So each 00:06:42.709 --> 00:06:45.089 department handles a different, pretty crucial 00:06:45.089 --> 00:06:46.990 part of the whole production. Yeah, exactly. 00:06:47.129 --> 00:06:50.649 So what's the big advantage, you think, of structuring 00:06:50.649 --> 00:06:53.110 this workflow into these distinct, almost like 00:06:53.110 --> 00:06:55.509 modular departments? Well, I think it effectively 00:06:55.509 --> 00:06:57.569 mirrors how a traditional studio works, right? 00:06:57.730 --> 00:07:00.149 Yeah. Which makes what could be really complex 00:07:00.149 --> 00:07:05.000 automation feel much more manageable and... Clear. 00:07:05.160 --> 00:07:08.240 So every great film, it starts with a great idea, 00:07:08.339 --> 00:07:10.459 right? And our automated studio is really no 00:07:10.459 --> 00:07:12.040 different. The whole process kicks off in the 00:07:12.040 --> 00:07:14.839 writer's room where, interestingly, two AI agents 00:07:14.839 --> 00:07:17.660 actually collaborate to generate this potentially 00:07:17.660 --> 00:07:20.600 endless supply of creative concepts. Yeah. The 00:07:20.600 --> 00:07:23.959 first agent acts as the ideas agent. You could 00:07:23.959 --> 00:07:25.720 call it the head writer or maybe the showrunner. 00:07:25.800 --> 00:07:28.740 Yeah. Its job is basically to come up with core 00:07:28.740 --> 00:07:31.139 concepts for your videos based on some general 00:07:31.139 --> 00:07:33.660 theme you give it, like, say, relaxing nature 00:07:33.660 --> 00:07:36.149 scenes or. daily historical facts, something 00:07:36.149 --> 00:07:38.170 like that. And it doesn't just output a single 00:07:38.170 --> 00:07:40.009 sentence, right? It creates a complete production 00:07:40.009 --> 00:07:41.949 brief, all structured in a way that the next 00:07:41.949 --> 00:07:44.149 AI can easily understand. Think of it like a 00:07:44.149 --> 00:07:46.370 really detailed, standardized recipe card for 00:07:46.370 --> 00:07:48.170 the video. So for example, it might generate 00:07:48.170 --> 00:07:50.910 crimson titanium shaped like a chestwork is cut 00:07:50.910 --> 00:07:52.790 cleanly in two, and then give a corresponding 00:07:52.790 --> 00:07:55.310 environment and sound prop too. And crucially, 00:07:55.370 --> 00:07:57.110 I think, once the head writer develops a core 00:07:57.110 --> 00:07:59.449 concept, it's automatically logged straight into 00:07:59.449 --> 00:08:02.069 a Google Sheet. This sounds simple, but it's 00:08:02.069 --> 00:08:05.110 a critical organizational step. It stops the 00:08:05.110 --> 00:08:07.589 AI from generating the same idea twice, lets 00:08:07.589 --> 00:08:09.649 you track the production status, and just gives 00:08:09.649 --> 00:08:11.790 you a complete overview of your entire content 00:08:11.790 --> 00:08:14.029 pipeline. Definitely. Then we have the prompts 00:08:14.029 --> 00:08:16.329 agent. This acts more like the screenwriting 00:08:16.329 --> 00:08:18.589 team. Its job is to take that single core idea 00:08:18.589 --> 00:08:21.990 and expand it out into multiple detailed scene 00:08:21.990 --> 00:08:24.230 descriptions or prompts to make the video dynamic. 00:08:24.649 --> 00:08:27.209 So from that titanium chest rook slice idea, 00:08:27.389 --> 00:08:29.629 it might generate three distinct scene prompts 00:08:29.629 --> 00:08:33.379 like... Macro shot, a dense orange burnished 00:08:33.379 --> 00:08:36.519 copper gear lies flat on a graphite slab. Or 00:08:36.519 --> 00:08:39.240 extreme close -up, an aluminum cube rests on 00:08:39.240 --> 00:08:42.080 frosted glass. You know, honestly, even with 00:08:42.080 --> 00:08:44.000 all these structured prompts, I still wrestle 00:08:44.000 --> 00:08:46.820 with prompt drift myself sometimes. Making sure 00:08:46.820 --> 00:08:49.080 the AI truly captures the nuance of the initial 00:08:49.080 --> 00:08:51.360 idea can still be a bit tricky. Interesting. 00:08:51.759 --> 00:08:54.740 So beyond just generating more raw ideas, what's 00:08:54.740 --> 00:08:56.940 the core benefit then of having two distinct 00:08:56.940 --> 00:08:59.200 AI agents working almost in tandem like this 00:08:59.200 --> 00:09:01.370 in the writer's room? They ensure you get both 00:09:01.370 --> 00:09:04.250 those high -level thematic ideas and the necessary 00:09:04.250 --> 00:09:07.389 detailed scene descriptions. Okay, so with the 00:09:07.389 --> 00:09:09.210 scripts written, production scheduled in our 00:09:09.210 --> 00:09:11.470 spreadsheet, it's time for the magic, right, 00:09:11.570 --> 00:09:13.490 on the soundstage. This is where the technical 00:09:13.490 --> 00:09:15.590 build happens, turning those words into actual 00:09:15.590 --> 00:09:18.570 visuals. Right. So first, we intelligently bundle 00:09:18.570 --> 00:09:20.909 those three scene prompts into a single dataset. 00:09:21.110 --> 00:09:23.789 It just makes it easier to handle. Then comes 00:09:23.789 --> 00:09:26.679 directing with C -Dance. This is really the heart 00:09:26.679 --> 00:09:29.460 of our studio. We use N8n to send our prompts 00:09:29.460 --> 00:09:32.460 over to wavespeed .ai, which acts kind of like 00:09:32.460 --> 00:09:35.139 an app store for AI models. It gives us super 00:09:35.139 --> 00:09:37.559 easy access to C -Dance. We configure it for, 00:09:37.659 --> 00:09:41.100 say, a 9 .16 aspect ratio, perfect for vertical 00:09:41.100 --> 00:09:43.799 videos like TikTok or YouTube Shorts, and tell 00:09:43.799 --> 00:09:46.779 it to make 10 -second clips. Gotcha. So after 00:09:46.779 --> 00:09:49.039 this step, our films are basically silent movies, 00:09:49.220 --> 00:09:51.100 right? No. Just the visuals. Correct. Totally 00:09:51.100 --> 00:09:54.110 silent. So next we introduce the sound department, 00:09:54.370 --> 00:09:56.690 which is powered by MM Audio via another service, 00:09:56.950 --> 00:10:01.190 Fel .ai. This AI is a master of Foley, you know, 00:10:01.210 --> 00:10:03.070 the art of creating realistic sound effects like 00:10:03.070 --> 00:10:05.789 footsteps or leaves rustling or glass breaking 00:10:05.789 --> 00:10:08.870 and just general sound design. It actually analyzes 00:10:08.870 --> 00:10:10.750 the video clips and automatically layers them 00:10:10.750 --> 00:10:12.950 with appropriate realistic sound effects. And 00:10:12.950 --> 00:10:15.149 the cost here, only about zero one cents per 00:10:15.149 --> 00:10:17.950 second of audio. We then briefly group the videos 00:10:17.950 --> 00:10:19.789 together again just to prepare them for the next 00:10:19.789 --> 00:10:22.279 stage. Okay. And finally, the editing bay. Yep, 00:10:22.320 --> 00:10:25.460 the final cut. For editing, we use FFmpeg, again 00:10:25.460 --> 00:10:29.379 via ale .ai. FFmpeg is this incredibly powerful 00:10:29.379 --> 00:10:31.759 open -source tool, kind of like the Swiss Army 00:10:31.759 --> 00:10:34.820 knife of video manipulation. It stitches our 00:10:34.820 --> 00:10:37.559 three 10 -second clips into one seamless 30 -second 00:10:37.559 --> 00:10:39.559 video and layers in that audio we just generated. 00:10:39.860 --> 00:10:43.139 The cost here is, well, it's essentially negligible. 00:10:43.220 --> 00:10:45.220 Less than a hundredth of a cent per calculated 00:10:45.220 --> 00:10:48.399 second. Basically free. And for smooth operation, 00:10:48.740 --> 00:10:51.700 the studio logistics are crucial. We have to 00:10:51.700 --> 00:10:53.779 build in these strategic pauses, maybe of a few 00:10:53.779 --> 00:10:55.759 minutes after video generation, a minute after 00:10:55.759 --> 00:10:58.639 sound is added, just to make sure each AI department 00:10:58.639 --> 00:11:01.139 has fully finished its complex task before the 00:11:01.139 --> 00:11:03.659 next step kicks off. It just keeps the whole 00:11:03.659 --> 00:11:05.779 assembly line running smoothly. Plus, of course, 00:11:05.840 --> 00:11:07.559 you'll set up your API keys for services like 00:11:07.559 --> 00:11:11.220 Wavespeed .ai and Foul .ai within N8n's secure 00:11:11.220 --> 00:11:13.580 credentials area. Right. Makes sense. And the 00:11:13.580 --> 00:11:15.679 very last step, then, to complete our automated 00:11:15.679 --> 00:11:18.309 studio workflow. Logging the final video link 00:11:18.309 --> 00:11:19.889 right back into our Google Sheets spreadsheet. 00:11:20.169 --> 00:11:22.289 Closing the loop. So how critical are those strategic 00:11:22.289 --> 00:11:24.389 pauses, those wait periods you mentioned, within 00:11:24.389 --> 00:11:26.950 this whole automated workflow? Oh, they're absolutely 00:11:26.950 --> 00:11:29.370 essential for seamless operation. They basically 00:11:29.370 --> 00:11:32.330 ensure each AI department fully finishes its 00:11:32.330 --> 00:11:35.590 task. Mid -role sponsor, read placeholder, content 00:11:35.590 --> 00:11:37.929 to be provided separately. Okay, so pulling it 00:11:37.929 --> 00:11:40.480 all together. What does this all mean? What's 00:11:40.480 --> 00:11:43.899 the final budget look like for our 30 second 00:11:43.899 --> 00:11:47.179 high quality video produced by this automated 00:11:47.179 --> 00:11:49.450 film studio? Let's just recalculate quickly. 00:11:49.649 --> 00:11:51.909 So clip generation, our director, that's three 00:11:51.909 --> 00:11:54.129 10 second clips at zero dollars or three cents 00:11:54.129 --> 00:11:56.769 per second. Total zero nine cents, nine cents. 00:11:57.169 --> 00:11:59.289 Sound generation from the sound department adds 00:11:59.289 --> 00:12:01.889 30 seconds of audio at zero one cents per second. 00:12:01.950 --> 00:12:03.450 So it's another dollars or three cents, three 00:12:03.450 --> 00:12:05.870 cents. And the video stitching by our editor, 00:12:05.990 --> 00:12:08.350 as we said, essentially negligible, less than 00:12:08.350 --> 00:12:11.190 a single cent. Two sec silence. So our total 00:12:11.190 --> 00:12:13.190 production cost comes out to approximately $0 00:12:13.190 --> 00:12:17.470 .13 per 30 second video. Less than the price 00:12:17.470 --> 00:12:19.990 of like a single piece of candy. Consider that 00:12:19.990 --> 00:12:22.590 for a second. When you compare this to the $22 00:12:22.590 --> 00:12:25.269 .50 it would cost with previous AI tools like 00:12:25.269 --> 00:12:28.269 VO3, the scale of this disruption just becomes 00:12:28.269 --> 00:12:29.789 crystal clear, doesn't it? You're not just saving 00:12:29.789 --> 00:12:31.809 a little bit of money here. You're really operating 00:12:31.809 --> 00:12:33.970 in a completely different economic reality. It 00:12:33.970 --> 00:12:36.830 really is the democratization of the moving image. 00:12:37.399 --> 00:12:39.580 Yeah. Seedense represents this huge paradigm 00:12:39.580 --> 00:12:42.379 shift, making high quality video content genuinely 00:12:42.379 --> 00:12:45.580 accessible to everyone on a scale that was, frankly, 00:12:45.740 --> 00:12:48.740 previously unimaginable. So what are the profound 00:12:48.740 --> 00:12:51.360 implications then for creators, for businesses? 00:12:51.419 --> 00:12:53.720 What is this unlock that was simply out of reach 00:12:53.720 --> 00:12:56.000 before? Well, it unlocks strategies previously 00:12:56.000 --> 00:12:58.639 reserved only for like. multi -million dollar 00:12:58.639 --> 00:13:01.080 media companies right you can now realistically 00:13:01.080 --> 00:13:04.679 create and post new unique video content daily 00:13:04.679 --> 00:13:07.360 maybe even hourly that's content at scale you 00:13:07.360 --> 00:13:10.299 can absolutely dominate niches by targeting very 00:13:10.299 --> 00:13:12.960 specific audiences with tailored videos all running 00:13:12.960 --> 00:13:15.720 on autopilot and maybe most powerfully you can 00:13:15.720 --> 00:13:18.580 perform massive a b testing imagine running dozens 00:13:18.580 --> 00:13:20.620 even hundreds of variations of a marketing video 00:13:20.620 --> 00:13:22.399 to find the perfect message without completely 00:13:22.399 --> 00:13:24.580 breaking the bank it sounds like a fantastic 00:13:24.580 --> 00:13:27.419 win for creators really democratizing things. 00:13:27.600 --> 00:13:30.690 But I wonder Does this also pose any challenges 00:13:30.690 --> 00:13:33.289 for, you know, the human element in video production? 00:13:33.509 --> 00:13:36.110 What happens to the roles of traditional videographers 00:13:36.110 --> 00:13:39.429 or editors or even concept artists when AI can 00:13:39.429 --> 00:13:42.230 handle so much so effectively at such an incredibly 00:13:42.230 --> 00:13:44.269 low cost? That's a that's a really important 00:13:44.269 --> 00:13:47.250 question. Yeah. And while Sedans currently has 00:13:47.250 --> 00:13:49.129 limitations, you know, like a lack of native 00:13:49.129 --> 00:13:52.129 voice generation, a cap on clip duration, those 00:13:52.129 --> 00:13:54.929 feel like temporary hurdles. Honestly, the overarching 00:13:54.929 --> 00:13:57.769 trend is just undeniable. AI model costs are 00:13:57.769 --> 00:13:59.909 plummeting. while their quality is soaring. And 00:13:59.909 --> 00:14:02.049 this puts immense pressure on competitors like 00:14:02.049 --> 00:14:04.789 Google and OpenAI to slash prices and improve 00:14:04.789 --> 00:14:07.850 quality. It's accelerating this race to the bottom 00:14:07.850 --> 00:14:10.049 for cost and to the top for quality. And that's 00:14:10.049 --> 00:14:12.029 a race where creators and consumers are ultimately 00:14:12.029 --> 00:14:15.190 the winners. The human element, I think it shifts. 00:14:15.330 --> 00:14:17.409 It shifts away from the manual labor maybe towards 00:14:17.409 --> 00:14:19.950 the strategic oversight, the truly creative direction, 00:14:20.110 --> 00:14:23.929 the initial spark. Beyond just the astounding 00:14:23.929 --> 00:14:26.610 cost savings, which are obviously massive, what 00:14:26.610 --> 00:14:28.490 do you think is the truly transformative impact 00:14:28.490 --> 00:14:31.210 of this technology on the very nature of content 00:14:31.210 --> 00:14:33.850 creation itself? I think it really boils down 00:14:33.850 --> 00:14:37.129 to unprecedented scale, hyper niche targeting 00:14:37.129 --> 00:14:40.750 and incredibly rapid iteration. OK, so to recap 00:14:40.750 --> 00:14:43.950 our deep dive today, high quality AI video production. 00:14:44.440 --> 00:14:46.580 It's no longer a luxury just for big studios. 00:14:46.779 --> 00:14:49.120 It's now genuinely accessible and affordable 00:14:49.120 --> 00:14:51.700 for virtually anyone, thanks to these innovations 00:14:51.700 --> 00:14:54.299 like CDANCE and automation platforms like NEN. 00:14:54.779 --> 00:14:56.899 Exactly. And this isn't just about saving money. 00:14:56.960 --> 00:14:59.340 It's really a complete redefinition of the creative 00:14:59.340 --> 00:15:01.659 and business possibilities. We're seeing the 00:15:01.659 --> 00:15:04.000 end of an era defined by those high costs and 00:15:04.000 --> 00:15:06.340 technical barriers and maybe the dawn of a new 00:15:06.340 --> 00:15:08.539 one defined by creativity, strategy and just 00:15:08.539 --> 00:15:10.980 raw accessibility. So here's a thought to leave 00:15:10.980 --> 00:15:13.639 you with. If the cost and the technical barriers 00:15:13.639 --> 00:15:16.440 to professional video creation are truly gone 00:15:16.440 --> 00:15:20.539 or rapidly disappearing. What entirely new forms 00:15:20.539 --> 00:15:22.879 of visual storytelling or marketing might emerge 00:15:22.879 --> 00:15:26.039 that we can't even imagine yet? Consider what 00:15:26.039 --> 00:15:28.039 personal or professional projects this might 00:15:28.039 --> 00:15:30.399 unlock for you. Thank you for joining this deep 00:15:30.399 --> 00:15:32.259 dive. Until next time, keep exploring.