The argument in one line.
Claude can be taught to replicate a specific studio's visual style — Vox's rough-paper documentary look — by feeding it reference scripts, a curated design board, and a manual animation breakdown, then using it to direct an image/video generator that executes what it plans.
Read if. Skip if.
READ IF YOU ARE…
- A video editor curious whether AI can actually replicate a specific studio's visual style, not just generic templates.
- Someone building a repeatable AI content pipeline who wants to see the exact reference-gathering and prompting process, not just the finished output.
- A creator evaluating Claude + Higgsfield as a way to produce animated explainer or documentary-style content without hand-animating in After Effects.
SKIP IF…
- You want a hands-on After Effects or Premiere animation tutorial — every visual here is AI-generated, not manually keyframed.
- You're expecting the actual Vox editor/design/animator skill files — the video only says they exist inside a paid course, not in this breakdown.
The full version, fast.
The creator shows how to make Claude imitate Vox's rough-paper documentary style from scratch: write a script, generate a voiceover in ElevenLabs, then train Claude on three separate skills — one that plans scenes from a script by studying 10 real Vox videos, one that reverse-engineers Vox's design system from ~100 Pinterest references, and one that breaks down Vox's five recurring animation assets (text, main object, background, secondary objects, camera move). Because Claude can't generate images or video itself, it directs Higgsfield through an MCP connector to actually produce the designs and animations. The full six-scene documentary cost about $20 in Higgsfield credits and was assembled in Premiere Pro with a voiceover and soundtrack.
Free for members
Chat with this breakdown — free.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →Where the time goes.
00:00 – 02:14
01 · Cold open: script and voiceover
Claude writes a 60-second Vox-style script covering facts, numbers, a graph, a historical event, and a human story, then the script goes into ElevenLabs to generate a cinematic AI voiceover.
02:14 – 04:20
02 · Training the Vox editor brain
Ten real Vox documentaries are transcribed via ElevenLabs speech-to-text and fed to Claude along with their scripts, so it learns how Vox editors plan scenes from a script. The output becomes a reusable 'Vox editor brain' skill.
04:20 – 06:38
03 · Training the Vox design brain
About 100 Vox-style reference images are collected from Pinterest into a Figma board, then Claude analyzes the exported PDF to reverse-engineer Vox's color, shape, texture, and layout system into a second skill. Higgsfield is connected via an MCP connector since Claude can plan images but not generate them.
06:38 – 08:35
04 · Training the Vox animator brain
The creator manually explains Vox's five recurring animation assets — text pop-ups, the main object (paper-unfold), a static background, subtly-animated secondary objects, and camera pans or zooms — and Claude turns the explanation into a third skill for directing Higgsfield's animations.
08:35 – 11:13
05 · Ultimate Editors 2.0 course pitch
Sponsor break for the creator's own editor-training platform, Ultimate Editors 2.0, and an upcoming AI editing fundamentals course.
11:13 – 12:44
06 · Step 1: Claude plans every scene
With all three skills loaded, Claude reads the uploaded script and uses the Vox editor brain to break it into six scenes, each with a design direction and reasoning tied to the specific line being spoken.
12:44 – 15:15
07 · Step 2 and 3: designs and animations
Claude uses the Vox design brain to generate Higgsfield mockups for every planned scene, then the Vox animator brain to turn those designs into full jittery, rough-paper Vox-style animations. The whole six-scene edit costs about $20 in Higgsfield credits.
15:15 – 16:13
08 · The finished Vox documentary
The finished 60-second AI-generated Vox-style documentary about shrinking attention spans plays in full, stitched together in Premiere Pro with the voiceover and a synthwave soundtrack.
16:13 – 17:15
09 · Closing take: AI raises the skill floor
The creator argues AI hasn't killed editing jobs — it killed the need to learn complex software, meaning editors who understand content psychology and design will out-compete those who don't. Closes on a final pitch for the AI editing fundamentals course.
Lines worth screenshotting.
- Claude can't generate images or video on its own, so the workflow pairs it with Higgsfield (via an MCP connector) as the execution layer while Claude stays the planning and design brain.
- The style-replication method breaks into three separate trained skills: one for planning scenes from a script, one for visual design, and one for animation — each trained on a different kind of reference material.
- Ten real Vox documentaries were transcribed via ElevenLabs speech-to-text and fed to Claude alongside their scripts so it could learn why editors chose each scene, not just what the scenes looked like.
- About 100 Vox-style images were pulled from Pinterest into a single Figma board, then exported as one PDF so Claude could analyze the whole design system in one pass.
- Every Vox scene breaks into five recurring animation assets: text (pop-up), a main object (paper-unfold or position pop-up), a static background, secondary objects (subtle static animation), and camera movement (zoom or pan).
- Because Claude has never watched Vox's actual animations, the animation training had to be done by manually explaining the five-asset system in plain language rather than showing it examples.
- The full six-scene, 60-second AI-generated documentary cost roughly $20 in Higgsfield credits, including testing.
- The creator's script brief to Claude specified a 60-second Vox documentary covering one subject across six settings, requiring facts, numbers, a graph, a historical event, and a simple human story.
- For voiceover, the creator chose ElevenLabs' 'Alex' business-book-narrator voice at medium speed specifically because it matched Vox's cinematic narration style.
- The creator argues AI hasn't eliminated editing jobs — it eliminated the need to learn complex software, shifting the competitive edge to editors who understand content psychology and design.
- Claude was prompted to package each trained skill (editor brain, design brain, animator brain) as a reusable Claude skill file, rather than a one-off session.
You can teach Claude any studio's exact visual style, then let it direct the render.
WHAT TO LEARN
Claude can't generate images or video, so the real skill here is turning curated references into three separate trained skills that plan, design, and direct an execution tool like Higgsfield.
01Cold open: script and voiceover
- Write the script and generate voiceover first — the script drives every downstream planning and design decision, so lock it before touching visuals.
02Training the Vox editor brain
- Speech-to-text turns any reference video into a script-with-timestamps you can feed an AI, which is how the creator taught Claude to reverse-engineer Vox's scene-planning logic from 10 real documentaries.
- Package each trained pass as a reusable skill file rather than a one-off prompt, so the training work compounds across future videos instead of being redone.
03Training the Vox design brain
- Pulling ~100 reference images into one board and exporting a single PDF lets an AI analyze an entire visual system in one pass instead of guessing from a handful of examples.
- Claude and an image/video generator split the job: Claude plans and directs, the generator (Higgsfield) executes — know which tool does which before you start.
04Training the Vox animator brain
- When the AI can't watch and learn a motion style directly, break it into a manually explained asset system instead — Vox's animations reduced to just five recurring parts (text, object, background, secondary objects, camera).
06Step 1: Claude plans every scene
- Once the skills are trained, planning becomes a single prompt: upload the script and let the editor-brain skill map every line to a scene direction.
07Step 2 and 3: designs and animations
- A full six-scene AI-animated documentary can run around $20 in generation credits, which sets a real cost baseline for anyone considering this as a production method.
08The finished Vox documentary
- Final assembly still happens in a normal editor (Premiere) — AI generates the pieces, a human still cuts them together with voiceover and music.
09Closing take: AI raises the skill floor
- AI removing the software-skill barrier doesn't remove the competitive gap — it just moves the gap to who understands content psychology and design well enough to direct the AI effectively.
Terms worth knowing.
- Higgsfield
- An AI platform that generates images, video, and audio; used here as the execution engine that produces the actual Vox-style designs and animations Claude plans.
- MCP connector
- A connection added inside Claude (via Settings → Connectors → Add custom connector) that lets Claude call an external tool — in this case, direct Higgsfield to generate assets.
- Claude skill
- A packaged set of instructions Claude can save and reuse across sessions — here, the creator turns each training pass (editor, design, animator) into its own downloadable skill.
- Vox style
- The rough-paper-cutout animation look used by the news outlet Vox in its explainer documentaries — textured backgrounds, choppy stop-motion-style movement, and a specific color/type system.
Things they pointed at.
01:36toolElevenLabs
04:40toolHiggsfield
04:25toolPinterest
04:30toolFigma
08:35productUltimate Editors 2.0
Lines you could clip.
16:13
“Creating this Vox edit taught me one thing about AI. It's a lot better when people actually know how to use it.”
16:30
“What I would be worried about is other editors learning this before you, creating better edits than you, and ultimately closing higher quality clients than you.”
16:40
“AI has done one thing to the editing space. It has killed the need to learn complex styles and programs.”
Word for word.
Read-along
Don't just watch it. Burn it in.
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
metaphoranalogy
00:00Claude can now officially create Vox style documentaries like this one completely from scratch. Everything you're seeing right now, all six of these Vox style animations were created entirely using AI. And even if you've never used AI to edit a video before, in this video, I'm going to show you everything you need to know to create Vox style documentaries like this using nothing but Claude.
00:20First, we're gonna start from zero and create the storyline and voice over using AI. Then I'll show you how we trained Claude to think like a real documentary editor and plan out animations. And finally, we'll use Claude and Higgs field to create all six of these stunning documentary animations from scratch with Sound Design as well.
00:36And by the end of this video, you're gonna be able to use Claude to create stunning documentaries like this completely from scratch even if you don't know a single thing about editing. And real quick, if you'd like to learn AI video editing from a to z, we're building a full AI editing fundamentals course inside Ultimate Editors.
00:51But more on that later in the video. Now before we create the crazy animations, we first need the storyline and voice over that guide our documentary. So let's quickly do that.
00:59Every documentary needs a good story, and typically Vox covers political or historical problems backed up by numbers, facts, and cool stories. What makes their videos special is that these numbers and stories are brought to life by . Now I know you clicked on this video to see how the Vox style is created with Claude, so I'm gonna keep this part quick.
01:19Here's how we created the storyline. Inside of Claude, I asked it for a sixty second Vox documentary script. It covers one subject across six different settings.
01:27It has to include facts, numbers, graphs, historical event, and a simple story, just like Vox. I led Cloud process, and it gave me this. The script was really high quality and definitely felt like a Vox video.
01:37And on top of that, it included a real variety of scenes, a section with numbers, section with a graph, a human example, and even a short story. Now that we had the perfect script it was time to generate the voiceover. Vox documentaries are known for the cinematic voiceovers it almost feels like a narrator telling you a story.
01:52To get that same experience I took my script and headed over to eleven Labs. Eleven Labs has a lot of AI features, but the one that we need is this one right here, the text to speech. It turns any script into a high quality audio recording of someone reading it.
02:05So I clicked on it, pasted in the script that Claude just made, then on the right I chose a voice that fit the documentary. I decided to go with Alex the business book narrator, set the speed to medium, and clicked generate.
02:16After some processing, it created this amazing voice over that actually sounded like a real documentary, which meant that now we had both the story and the voice over, it was time to start creating the Vox documentary edit using AI. But before Claude can create a stunning Vox style edit, we first need to train it on the Vox style.
02:33You see, out of the box, Claude doesn't know how to edit a Vox documentary, but that's not a problem because we can teach it. We already have the script of our story, so the first thing Claude needs to learn is how to look at a script and plan out documentary animations and scenes based on it. Here's what I did.
02:48First, I downloaded 10 Vox documentaries that I really liked. All of them full of animations that are built around what's being said in the script. Then I headed into Eleven Labs and clicked on the speech to text function.
03:00This transcribed any video and gives you its script. I clicked transcribe, select the YouTube tab, and pasted the URL of each video.
03:07After processing, we get this, the full script of the video with timestamps exactly what's being said and when it's being said. Then I click on export, export as a JSON, and download. We now have 10 documentaries with all their animations, and we have their exact scripts with what's said and when it's said.
03:23So I fed all of this into Claude and told it to study these documentaries and their scripts, analyze what's being said, look at the video, and analyze what's being shown, and deduce why. Claude processed all 10 documentaries and created what we call . It's a complete guide on how the editors at Vox think when coming up with the scenes they wanna show.
03:42And you're gonna see just how important this skill is later in the video when we create the six sunning animations because it lets us feed Claude any script, and it automatically plans every scene for our documentary exactly how Vox does it. The best part is if you also want to create a Vox documentary using AI, you don't actually have to train Claude like I did.
03:59You see, I asked it to turn everything into a skill that you guys can download and tell Claude to save. And every skill that we create in this video is gonna be available inside my AI editing fundamentals course in Ultimate Editors when it goes live. If you're an editor that wants to learn how to edit all kinds of amazing videos with AI, check out Ultimate Editors.
04:16The link's gonna be down below. So now Claude can take a script and plan animations. The next step is turning that plan into Vox style designs, which means we now need to teach Claude how to design scenes just like Vox does.
04:27How on earth are we going to do that? Well, when you look at Vox scenes, they have a lot in common. They use a certain design language, common fonts, common colors, and common styles.
04:37We just need Claude to look at these and analyze the design system that Vox uses. Here's how. First, I needed to collect references of the Vox design.
04:44So I head into Pinterest since I can find a lot of designs there and search the term Vox documentary animation design. I spend around fifteen minutes collecting a 100 different high quality Vox designs that all fit the style, and I place them all in this Figma file like so. So now we have a wall of Vox designs that Claude can study.
05:01Then I took the Figma PDF of this inspiration wall, uploaded it to Claude, and prompted it to study every one of these designs and find the patterns and design systems that make up the Vox styles. The colors, shapes, text, objects, macro designs, micro details, and textures. And after letting it process, it came up with this in-depth PDF called the Vox Design Guidelines, which broke down exactly how Vox designs all of their scenes.
05:24Honestly, this was the most important part of creating the six stunning Vox animations as you're about to see in a bit. And of course, I asked it to turn all of this information into a second skill for you guys, the Vox design brain. There's just one thing to keep in mind.
05:37Claude knows the Vox style, but Claude can't actually generate images. This is why we use a platform like Higgs Field which can generate images, video, and audio. Since Claude knows the design and Higgs Field can generate the images, we just need to connect Claude to Higgs Field where Claude teaches it the design and Higgs Field executes the image.
05:54If you haven't connected Higgs Field to Claude yet, it's super simple. Go to this website, copy the MCP link, then head into Claude, click customize, connectors, add add custom connector, and paste the link here. Now just look at how far Claude has come.
06:06It can now look at a script and plan out every animation that needs to be on the screen and when using the Vox editor brain, and then it can take that plan and turn it into high quality Vox designs that fit the branding using the Vox design brain. Just stay tuned to see how easy these skills make it to create the six amazing Vox style animations that I'm gonna be making later in the video.
06:25But now this means that it's time for the final step turning these designs into high quality Vox style animations. How on earth are we going to get Claude to do that? Take a look.
06:34Vox animations have a lot in common. The choppy frame rate, the movement, the simple camera movement.
06:40This means that just like design, Vox uses a certain system for animating their scenes. But we can't get Claude to watch the animations, so what I did was go through multiple animation scenes from Vox and manually explain to Claude everything that's happening. Every Vox scene breaks down into five major assets.
06:56There's the text which usually has this minimal text pop up animation. Then there's the main object which typically has a paper unfolding animation or some kind of position pop up. Then there's the background which doesn't really animate.
07:05Then there's the secondary objects. These are just here to add more design, and they usually have some kind of subtle static animation. And finally, there's the camera movement, which sometimes has a zoom in or sometimes a camera pan around the scene.
07:16After explaining all five of these to Claude, it understood the assignment and built the third skill, the Vox animator brain. This was the vital skill needed to create the six stunning Vox animations at the end of the video. I'll show you how we built them in just a second.
07:30And just like the designs, Claude can't actually create animations, but Higgs Field can. So Claude is the brain that understands how Vox animations are created, and it directs Higgs Field to create the animations themselves. So with all this training wrapped up, Claude can now do three huge things.
07:44First, it can look at any script and use the Vox editor brain to plan out the exact scenes we need to create for the viewer. Second, it can take this plan and use the Vox design brain to create Vox style designs that look amazing and fit the branding. And third, it can take those designs and use the Vox animator brain to create animations that look exactly like the ones Vox creates in their videos, which now officially meant that Claude could theoretically create a full Vox documentary if we give it the script.
08:10And that is exactly what we did. But before we put it to the test and create the six stunning Vox animations that you saw at the beginning of the video, you might either be an editor that wants to learn how to use AI in your workflow or even just a beginner editor that just got into video editing and feels overwhelmed with how much there is to learn.
08:26Let me tell you, this is completely normal and we all had to go through it. I recently interviewed our students and they all mentioned that learning editing on YouTube is overwhelming and unstructured. That's exactly why I decided to build Ultimate Editors two point o.
08:39It's the best platform for beginner editors to master the viral editing styles of 2026 and make $2,000 per month in ninety days. But now we added something even better. First, you're going to get access to the Premiere Pro and After Effects master classes to learn the editing programs step by step.
08:53Then you're going to get access to the viral editing style breakdowns of 2026. The cinematic style, the minimal animation style, the sass motion design style, and the 30 plus viral animation breakdowns. Then on top of all of this, you're even gonna get access to the 2,000 per month editor blueprint.
09:08So you're not only gonna learn the editing programs, master the viral editing styles of 2026, but you'll even learn how to make $2,000 a month as a video editor. But that's not it. With AI advancing fast, we're all struggling to keep up and learn how we can incorporate AI into our workflow to make us better editors.
09:24That's why I created the AI editing fundamentals course where I'll teach you everything about editing amazing AI videos like this Vox documentary from scratch. And this course is launching in about two weeks. If this sounds interesting and you'd love a step by step course on video editing, the viral editing styles, and AI video editing, it's all available inside Ultimate Editors two point zero.
09:44The link is gonna be down below. Click the link, and we'll see you guys inside. But now that Claude was trained and ready, it was time to put it to the test and create the six stunning Vox animations that you saw at the beginning of the video.
09:55With Claude fully trained and all the skills inside, here's how we create a Vox documentary using nothing but Claude. Before I start, I first prompt Claude to load up all three Vox documentary skills, the editor brain, the design brain, and the animator brain. Then I tell it I'm going to ask it to edit a Vox documentary for me.
10:12It all starts with step one, planning out the edit. I upload my script and I prompt Claude to read through it and use the Vox editor brain to plan out every single scene for the video just like a Vox documentary. After some processing, it thinks exactly like a Vox editor and creates a full plan of every scene it's going to create.
10:28Here's exactly how it works. Every phrase that's being said in the script gets a breakdown of what the big focus should be, and based on this, a general guide on what the design needs to include. My documentary starts with a phrase about how in 2004 the average person could focus for two and a half minutes whereas now it's only forty seven seconds.
10:43So Claude planned to introduce a year counter then transition to a character on the left with text on the right highlighting it and then to finish it off with a clock and text overlapping it. This was literally perfect. The second scene was about how this is happening from Tokyo to Toronto across the globe.
10:57So Claude planned a map animation of both cities then transitions to a global map with red circles mapping out each city. The third scene was about attention span dropping year by year exactly in 2007 when the iPhone dropped. So Claude planned a graph with a circle around the year 2007, then transitioning into a micro shot of the graph to emphasize this.
11:16My fourth scene introduced the average worker, how they check emails a lot, and how the reason is that the apps are designed to do that. So planned an introductory scene revealing the character in the middle, then two object scenes showing the character in a MacBook, and finally transitions into a big circle that closes the whole scene.
11:31My fifth scene talks about how character's workday is divided into percentages. So Claude planned a circle graph around the character highlighting every percentages needed. Finally, my sixth scene was a concluding scene with a quote.
11:42So Claude planned to simply write it on the screen with some extra design assets. This plan was more than perfect. Claude thought exactly how the Vox documentary editors do and created scenes based not only on what's being said in the script, but also on how we want the viewer to feel throughout the video.
11:57Now with our plan created, it's time to bring it to life. This brings us to step two, creating the Vox style designs. I prompt Claude to use the Vox Design Brain to design every scene it planned exactly how Vox would.
12:08The rough paper, the textures, and the fonts. I tell it to use a blue teal accent color because I think it would look good and use Higgs field to create the design mockups. It processes for a few minutes and then it starts sending me designs for every scene from the table.
12:21For my first scene, it gave me this design for the year counter, this design for the character and text, and this design for the clock and text. All of them matched the Vox style perfectly which was exactly what we needed. For my second scene, it generated this map image of Tokyo, this map of Toronto, and this finishing global map scene.
12:36This could have been done much better but it still did an amazing job. For my third scene, it created this vox style graph with the highlight around the year 2007. Then this micro shot of the graph with the year circled and it looked amazing.
12:47For my fourth scene it created this character introduction with highlighted text, this side by side setup of the MacBook and this concluding circle closing scene and this was perfect. For my fifth scene it came up with this simple character and pie graph diagonal design and I honestly loved it. And finally for the sixth scene, it created this text focused design which was awesome.
13:04Using the Vox design brain that we taught it, Claude could create stunning designs that looked exactly like Vox and honestly this was just mind blowing. Remember if you guys also want access to these skills and wanna learn how to edit videos like these using AI, the link to Ultimate Editor is gonna be down below. You're definitely gonna wanna be inside by the time that we release our AI editing fundamentals course.
13:23Now that every scene has its design, it's finally time to turn them into full animations and create the Vox style documentary, which brings us to the final step. Step number three, creating the Vox style animations. Here, I prompted Claude to use the Vox animator brain to turn every single one of those scenes with its creative direction into a Vox style animation using c dance two point zero in Higgs field.
13:43I told it to make sure to use the rough choppy jittery Vox look on all the animations and for each scene to transition between the designs to make the edit flow a lot better. Now this is the step that costs the most money in credits, But don't worry. This whole edit from zero cost me around $20 in Hicksfield credits, and that's with testing.
14:00Claude first processed the three designs for scene one and created this amazing year pop up character reveal and text highlight that transitions into the clock scene at the end. This looked like something straight out of a Vox documentary. For scene two, it created this map highlight that transitions to a different city and zooms out into the global map while highlighting the two big countries with red circles.
14:18It could still use some work, but it looked awesome. For scene three, it created this graph draw out scene with a line dropped, then slowly transitions into the micro shot of the graph at 2,007. The transition was rough, but the overall animation looked amazing.
14:31For scene four, it started with a paper crumble reveal of the character in the middle, then transitioned to the two object scene with the MacBook before zooming out and closing it all in one circle. Really solid flow. For scene five, it created the reveal on the left side with the circle graph, drawing out on the right side, which all looked great.
14:48And finally, for scene six, it finished it off by writing the text out, highlighting it, and having a hand icon pop in from the bottom of the screen. This was crazy. Using the Vox animator brain, Claude was able to nail the animation style down to a tee.
15:00And with all six animations created, all I did was download them, import them into Premiere Pro with the voice over, align each one where it belonged, and add a mysterious synthwave soundtrack that felt like Vox. And with all this complete, this is what our final edit looks like. In 2004,
15:15the average person could focus on one thing for two and a half minutes. Today, forty seven seconds.
15:23This isn't happening in one place. From Tokyo to Toronto, every country that got smartphones got the same problem, and researchers can map the moment it arrived.
15:35Average attention span measured year by year doesn't decline gently. It falls off a cliff right here, 2007, the year the iPhone shipped.
15:46Meet the average worker. She checks emails 74 times a day, not because she wants to, because every app is engineered to pull her hand back.
15:57Here's her workday, 60% fragments shorter than ten minutes. Only 3%.
16:04This sliver is deep focus. So the real question isn't how we lost focus. It's who profits from never giving it back.
16:13Ladies, gentlemen, creating this Vox edit taught me one thing about AI. It's a lot better when people actually know how to use it. If you're concerned about AI ending video editing jobs, I wouldn't be worried about that, and I'll make a separate video on why.
16:26What I would be worried about is other editors learning this before you, creating better edits than you, and ultimately closing higher quality clients than you. Because AI has done one thing to the editing space. It has killed the need to learn complex styles and programs.
16:40Instead, now you just need to understand content psychology, which is how good content's edited. You need to understand good design, which is what makes a good scene design and what doesn't.
16:49And, ultimately, you need to understand how to edit videos with AI. That's exactly why I decided to create the a to z AI editing fundamentals course in Ultimate Editors, where I'm gonna be teaching you everything about AI video editing, from the AI programs to the training to the skills. On top of this, our editing fundamentals course also teaches you how to design good scenes, and our Viral Talking Head masterclass teaches you advanced editing psychology.
17:11Thanks a lot for watching all the way till the end of the video. Stay tuned for our next banger, and we'll see you.
The Hook
The bait, then the rug-pull.
Claude can't generate a single image or frame of video on its own — so this creator taught it to direct one that can. By feeding Claude ten real Vox documentaries, a Pinterest board of design references, and a manual breakdown of Vox's five animation assets, he turns it into a three-skill production brain that plans, designs, and animates a full documentary through Higgsfield.
Named ideas worth stealing.
02:14model
The three-skill training pipeline
- Vox editor brain — plans scenes from a script by studying 10 real documentaries and their transcripts
- Vox design brain — reverse-engineers the visual system from ~100 curated reference images
- Vox animator brain — directs Higgsfield's animation of each design using a manually taught 5-asset breakdown
The three-step method used to teach Claude an entire studio's editorial process: how to plan a scene, how to design it, and how to animate it.
Steal forReplicating any specific creator or studio's content style with Claude paired with an image/video generator.
06:39list
The 5 Vox scene assets
- Text — minimal pop-up animation
- Main object — paper-unfold or position pop-up
- Background — stays static
- Secondary objects — subtle static animation, adds design detail
- Camera movement — zoom-in or pan around the scene
The recurring animation-asset breakdown the creator manually taught Claude so it could direct Higgsfield's Vox-style scene animations.
Steal forAny AI-driven animation pipeline that needs to replicate a specific studio's motion-design system.
How they asked for the click.
VERBAL ASK
08:35product
“check out Ultimate Editors, the link's gonna be down below”
Sponsor break placed right after the three skills are established, tying the pitch directly to what was just demonstrated; repeated again at the very end as the sign-off CTA.
FROM THE DESCRIPTION
PRIMARY CTAWhere the creator wants you to go next.
OTHER LINKSAlso linked in the description.
Visual structure at a glance.
