14:06Joseph | Video Editing · Essay
How I'd Learn Video Editing in 2026 (Before AI Replaces You)
A 30-day roadmap from zero to paid AI video editor, built on the idea that editors get paid for vision, not for knowing Premiere Pro.
November 12thThe argument in one line.
Claude can edit a premium talking head video only if you first teach it the style through references and a breakdown, and then plan the edit yourself, because humans still plan better than AI.
Read if. Skip if.
READ IF YOU ARE…
- A video editor who wants to understand how the premium minimal talking head style is built: palette, fonts, captions, B-roll and full-screen animation scenes.
- Someone already using Claude or Claude Code who wants a repeatable method to teach it any visual style by feeding it references and a rulebook.
- A beginner editor trying to work out where AI editing fits into a short-form workflow and what the human still has to do.
SKIP IF…
- You want a click-by-click Claude setup tutorial. The training and editing steps are shown as a demo, and the full skill packs are sold separately.
- You edit long-form or cinematic content only. Everything here is about vertical 9x16 talking head edits.
- You are looking for neutral advice. The last third is a sales pitch for a $299 course that closes November 15.
The full version, fast.
Premium talking head edits come down to four principles: a four-color palette pulled from the footage with two font families, A-roll with hook text and six-to-eight-word subtitles, stylized B-roll, and four kinds of full-screen animation. To get Claude to produce them, collect 10 reference edits, screenshot every scene into a Figma inspiration board, and have Claude study both and write a skill with 235 broken-down animations. Then you plan each scene yourself and let Claude build, render and sound-design the edit.
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 – 00:58
01 · Promise and roadmap
Joseph promises an AI-edited viral Reel using only a Claude subscription and lays out three parts: four principles, training Claude, then editing a raw video.
00:58 – 06:20
02 · Four principles of the style
Colors and fonts, A-roll and subtitles, B-roll scenes, and four kinds of full-screen animation.
06:20 – 08:18
03 · Course pitch: Ultimate Editors 2.0
Why understanding the style matters, then a first pitch for the five masterclasses, AI editing fundamentals and the Joseph skills.
08:18 – 10:14
04 · Collect references and build the Figma board
Gather 10 reference edits, screenshot every scene into Figma, and group the scenes into hook text, body text, B-roll and full-screen animation.
10:14 – 13:00
05 · Train Claude and inspect the skill
The study prompt, the 235 broken-down animations, and the four-part skill file Claude produces.
13:00 – 13:53
06 · Skill packs and second pitch
Joseph packages each trained style into a skill pack and sells five of them inside Ultimate Editors.
13:53 – 15:21
07 · Prepare the raw clip
On Fable 5.2 at high effort, Claude transcribes a Huberman clip, cuts silences to 60 seconds and crops to 9x16 tracking the speaker's nose.
15:21 – 17:33
08 · Plan every scene by hand
Joseph writes the plan in a script-and-boxes document: palette, hook text, speaker intro, B-roll with selector boxes, molecule animations and A-roll transitions.
17:34 – 19:53
09 · Claude builds, renders and sound designs
Claude collects assets with the Chrome extension, builds captions and scenes in about 45 minutes, renders 1080p, then adds sound effects and a chosen music track.
19:53 – 21:02
10 · Final result
The finished Huberman edit is played in full.
21:02 – 22:05
11 · Closing pitch
The one big takeaway for editors, then the Ultimate Editors 2.0 offer: $299 lifetime access, closing November 15.
Lines worth screenshotting.
- Pick one accent color from your footage, then derive a bright version, a dark version, and a white and black tinted toward it. Those four colors cover every background, element and text.
- Premium talking head edits use two font families: a regular sans for body text and a special style for key words, with thin weights of 300 to 400 and special words at 600 to 700.
- In a hook scene the speaker sits in the upper half, the most attractive phrase goes in a special font in the center, and a name-and-title callout sits bottom left.
- Subtitle six to eight words per scene, pick one special word per scene, and align the rest of the words around it.
- Full-screen animations fall into four kinds: two-object alignments, graph and number scenes, real elements snapped into animations, and one-object-centered scenes.
- If you don't understand how good edits are made, you will only use AI to make more bad edits faster.
- To train Claude on a style, collect at least 10 full reference edits from a creator who already does it well.
- Screenshot every well-designed scene from the references into a Figma board, then group them by pattern such as hook text, body text, B-roll and full-screen animation.
- Claude broke 10 reference videos into 235 animations, each with a design still, a keyframe sheet and a written breakdown, and grouped them into 14 scene categories.
- A style skill has four parts: a raw breakdown file, an inspiration guide, a lookbook of palettes and fonts, and the skill file that ties them together.
- Track the speaker's nose and keep it on the vertical center to turn a horizontal podcast clip into a clean 9x16 crop.
- The creator's rule from editing with AI: humans are still better at planning what the viewer should see, so he writes the scene plan himself and hands it to Claude.
- Transitional statements should stay as A-roll with subtitles to carry the viewer into the next visual scene.
- Claude built every animation for the demo edit in about 45 minutes, then rendered the 1080p video and synced it to the voices.
- Sound design can be trained as its own workflow: audit the video, place effects on timing and volume, then preview 10 music tracks and pick one.
Teach the style first, then plan it yourself
WHAT TO LEARN
AI edits only as well as the style knowledge and scene plan you give it, so the real work is building the references, the rulebook and the human-written plan.
02Four principles of the style
- Pull one accent color from the footage and derive a bright shade, a dark shade, and a tinted white and black. Four colors are enough to unify backgrounds, elements and text.
- Use two font families, a regular sans plus a special style for key words, with thin weights for body text and heavier weights for the one special word per scene.
- Subtitle six to eight words per scene, and match each scene type to what the speaker is saying: single object, numbers, a real experience, or a transition that stays on A-roll.
04Collect references and build the Figma board
- Understand how good edits are built before using AI, otherwise the AI just produces bad edits faster.
- Collect at least 10 strong references and screenshot every scene onto a Figma board, grouped by pattern, so the breakdown covers hook text, body text, B-roll and full-screen animation.
05Train Claude and inspect the skill
- A style skill needs four layers: a raw study of the references, a categorized inspiration guide, a lookbook of colors and fonts, and a skill file that applies them.
07Prepare the raw clip
- Prepare raw footage by transcribing it, cutting silences and cropping to 9x16 with the speaker's face tracked on the vertical center before any design work starts.
08Plan every scene by hand
- Write the scene plan by hand, line by line against the script, because humans still decide better than AI what the viewer needs to see and feel at each moment.
09Claude builds, renders and sound designs
- Let the AI do the production loop of asset collection, scene builds, review sheets, rendering and sound design, and review each stage before moving on.
Terms worth knowing.
- A-roll
- The main footage of the person talking to camera, as opposed to cutaway visuals laid over it.
- B-roll
- Supporting footage or images shown over or instead of the speaker to illustrate what is being said.
- Lower third
- A graphic in the lower part of the frame, here a name and job title callout that introduces the speaker.
- Selector box
- A tracked rectangle drawn around an element in a clip to point the viewer at it, often paired with an arrow and a label.
- Skill file
- A packaged set of instructions and reference material that Claude loads to perform a task in a consistent way, here editing in one specific style.
- Lookbook
- A document listing the color palettes and font styles of a design style and how they should be applied.
- Keyframe sheet
- A breakdown of an animation's key positions over time, used to reproduce its motion.
Things they pointed at.
00:00toolFable 5.2
00:00toolClaude
02:00toolPP Neue Montreal
02:00toolUrbanist
08:47channelDope Motion Zone
08:47toolFigma
13:53channelChris Williamson and Andrew Huberman podcast
18:06toolChrome extension for Claude
00:29productUltimate Editors 2.0
Lines you could clip.
06:40
“If you don't know how to make good edits, you're simply going to be using AI to make more bad edits.”
15:21
“One thing that I learned by editing with AI is humans are still better at planning how the edit should go and what the viewer needs to see on the screen.”
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.
In this video, I'm going to show you how I used Fable 5 .2 to edit this stunning viral Instagram Reel using nothing but AI. And no, you don't need Hicksfield. And no, you don't need to know anything complex about video editing.
All you need is the Cloud subscription. In this video, I'll first show you the four principles to creating viral talking head edits. Then I'll show you how you can train Cloud on any style and create a short form editor skill yourself.
And finally, I'll show you how to use this skill to turn a raw video into this premium edit inside of Cloud. And by the end of this video, you'll know everything there is creating stunning viral edits like these using cloud if you would like to learn how to edit stunning viral videos in 2026 like these we have step -by -step beginner -friendly master classes in ultimate editors 2 .0 but you won't be able to buy access to them after november 15.
so click the link down below before it closes down forever but with all this being said before we start creating viral edits using ai let's first start by understanding how a viral talking head video is even edited the better we can understand how good talking head videos are edited the better we can guide the ai to create a stunning edit if we take this premium editing style and break it down we're gonna notice it's really just made up of four major principles these four principles are the secret to creating stunning viral talking head videos just like these the first principle is the colors and the fonts the most special thing about this editing style is the complexity of the color theory and the fonts it's what really makes these edits feel so premium so after breaking down all of these videos i'm going to try and explain it to you in the simplest way in order to pick the colors you start by looking at your raw footage and the first thing you do is
collect your accent color sources look at the scene and ask yourself what colors can i see the speaker shirt the lamp in the background the blue window or the color of the background these are all accent colors that you can use now you need to pick one of those accent colors for me i'm choosing the yellow lamp and then you need to extract a vibrant bright version of that accent color which would be the gold and a dark rich version of it which would be the orange then you need to grab a white and a black and tint both of them to the shade of your accent color which in my case means making both of them more warm these four colors will be used throughout your edit for the backgrounds, the elements, and the texts based on its color theme.
Now, this is just scratching the surface, but what about the fonts? For the fonts, there are two main families always being used. The first one is the regular text, which is usually a sans font like PP New Montreal or Urbanist, and the second is for special words, and it could be anything between a script font, a serif font, or in my favorite case, just a bolder version of the sans font.
As for the font properties, throughout the A -roll and the full -screen animations, for your thin text, we're using a weight of 300 to 400, and for any special words we're using a weight of 600 to 700. Now I know these sound very basic but when you see how we're going to be using them throughout the edit you're going to notice just how important they are.
The second principle is how we handle the A -roll and the subtitles. The most common type of scene in this editing style are the text scenes. So how do we build them?
First there's a special way of editing hook animations in this talking head style. You always place your character in the upper half. This helps the viewer connect to the speaker and see who it is.
Then in the lower half this area here is reserved for your hook text. Now the way your hook text is built is you see simply look at the hook that you're saying and identify the phrase that is most attractive.
It could be a negative word or a big question. Then you place that phrase in a special font in the center. This is going to help the viewer instantly read the statement that makes him hook to the video.
After that, you can place the rest of the words using your regular font around your main text. To finalize your hook, there's one last element that needs to be added. In the bottom left of the screen, you add a callout introducing the speaker, a main title saying his name, and a subtitle describing his position.
There are many designs that you could choose from here. This is how you create good hooks in this editing style.
But what about the text for the rest of the edit? It's also very simple. You simply subtitle six to eight words for every scene or one phrase.
Then for each scene, you find a special word, place it in a special font and align the sands around it in a way that you find fitting. There are many ways to creatively subtitle your video, but if you use this strategy with the right colors and the right fonts, you're going to have very nice subtitles. The third principle is the B -roll scenes.
These are where the style really shines. Oftentimes in the video, you're going to want to show B -roll or images and this style allows you to do that in a very artistic way.
You can show the full clip, have a rough texture effect on it to emphasize it, a darken and blur effect on the edges, and a cinematic text design in the middle to deliver the message. Or you could swap the cinematic text for an arrow and text design pointing to the subject. If there is a certain element you want to highlight, you can even use a tracked selector box around the element.
And if you just want to showcase text, you can simply use leading lines with a camera pan animation to introduce the text and lead out of the scene. As for the images, these are even more artistic. You can create a side -by -side layout where you showcase the subject, on one side of the screen in an image, and then you lay out the cinematic text on the other side, sometimes even behind the subject to describe the scene.
This always works if you wanna show both the viewer and an image, but also wanna explain. The second way you could go about it is incorporating the design into the image. If your main subject is in the center, you can add cinematic text behind the subject or even add graphic lines and textures around your subject to emphasize them.
This is really useful if your subject is tied to the center of the screen. So anytime you wanna show B -roll and images, you can use these rules to stylize them well. Finally, this brings us to the Big fourth principle, the full screen animations.
These are the real bread and butter of the editing style. The first major kind of scene setup that we see is two object alignments. In side -by -side alignments, the object is placed on one side, typically with a casted shadow, and the other side is used to align cinematic text.
Then finally, you're going to find blurred and faded designs on the edges that make the scene more cinematic. You can also create upper and lower alignment where you can find cinematic text in the top and a design element that takes up the lower third into the center with filler elements around it. These scenes are used anytime the subject talks about one scene.
singular thing and you want the viewer to focus on that. The second kind that you're going to find is the graphs and number scenes. In these, you can have graphs in the center of your screen with text around it describing it, or you can have a bar graph in the lower half of the screen with clean text up top describing it.
Or you can create patterns of buildups using illustrations. These are great anytime your subject speaks about increases or decreases in numbers. The third kind of scene you're going to find is real elements incorporated into animations.
In here, you can have a scene with real images or videos snapping to a background with this clean stroke effect around it. You can even split them apart in the layout and place good text in the center or even lay them out horizontally and place a subject under them with text above.
These are really good anytime your subject talks about an experience or something that the viewer should see and understand. And the fourth kind of scene that you're going to find is the one object centered scenes where you can showcase a singular element in the middle with a casted shadow like this and depth elements on the edges like so or you can have an illustration take up the center of the screen in a clean way or simply place an object showcase in a realistic 3D scene like this one with text.
And these are always good if you only want the viewer to see a singular object these four kinds of scenes really stand out when good design is used in them but you can use them accordingly depending on what your subject is saying throughout the edit and these four principles that you just learned the color fonts the captions the bureau and images and the full screen animations all come together to form the stunning editing style that you know and love now you might be wondering if we're going to be editing this video with ai why does it matter if i understand how the style is made or not because simply put if you don't know how to make good edits you're simply going to be using ai to make more bad edits this This applies to design, content psychology, and creating stunning scenes.
Now, if you're new to video editing, perhaps you've noticed the new editing styles coming up, AI video editing rising, and now AI video editing reaching new levels like never before. Maybe you feel confused or overwhelmed and you're not sure where you can learn all this stuff. This is why we built Ultimate Editors 2 .0, the step -by -step program to learn viral editing, master AI video editing, and make $2 ,000 per month as a video editor.
And this program is closing down forever in just a few days. So listen up while you still can. Inside, you're gonna get five step -by -step masterclasses the viral editing styles of 2026 cinematic short films minimal animations sass trailers viral animation breakdowns and even the viral talking head editing style we're teaching you in this video we're even going to teach you the design fundamentals and content psychology so you fully understand how good videos are made as for ai editing our ai editing fundamentals course will walk you through the tools claude remotion and hicksfield from a to z even if you've never seen them before then we're going to show you how to edit iman gaji style videos fox documentaries and viral short form edits using ai and here is the best part i've been editing editing my short form content and these very long form videos using AI for the last month or so.
So I decided to create the Joseph AI editing masterclass where I give you my Joseph short form editing skills that I use to edit viral videos for my Instagram and my Joseph long form editing skills that edit the very video you're watching right now. And even step -by -step courses teaching you how to use it. Again, there's so much in Ultimate Editors 2 .0, but it closes down forever in just a few days.
So while it's still available, if you're a beginner editor that wants to learn how to edit viral videos in 2026 and get a head start on AI video editing, click the link down below and join Ultimate Editors 2 .0. But now that we've broken down all the principles that make up this premium talking head editing style, let's go ahead and see how we can train Claude to create stunning edits in this style.
Now, I've done this over and over to multiple different editing styles. And honestly, the best workflow that I found to train Claude on any new editing style is this one that I'm gonna show you right now. Step one, you need to collect references.
You need to find full edits made in the style that you like and collect at least 10 good references. For our style, I found this guy named Dope Motion Zone. drop him a follow by the way, he's a sick editor.
And I went through his personal brand and selected the 10 best videos that I found. I downloaded them and placed them in a reference folder like so. Step two is to then turn your reference style into an inspiration board.
This is where you need to load up a program like Figma and load up the reference videos you just downloaded. Then as you're watching through the videos, every single time you see a well -designed scene, you screenshot it and you drag it onto the Figma board. You do this for text scenes, B -roll scenes, and even full animations.
And you do this for every single scene in those 10 stunning videos. By the end, you should have a full inspiration board of what your editing style is made up of.
And this is where the magic happens. Step three is now you need to break down the editing style. You begin by looking at the inspiration board and searching for patterns.
For example, I noticed that all these scenes had one thing in common. They were simply A -roll and subtitles. So I collected them here.
Then I started by breaking them down. I noticed the batch is how the text looks in the hook. So I added a section called hook text design.
I drag all the hook text references here on the left. And I take one that I like and break down what it's made up of. The text, the blur, the lower thirds introduction.
And I add a quick written breakdown of how it's created next i noticed that the rest of the inspiration is all about how the text is made in the body so i dragged that down into the body text design take a few special ones break them down too and write a detailed description on how it's built then we looked at the inspiration board and noticed that there's a lot of sections with b -roll so we collected them and then we learned that there's four kinds of bureau scenes with footage and two kinds of bureau scenes with images and finally we collected the rest of the inspiration into the full screen animation section where we broke down all scenes with two objects we broke down all graphs and numbers all real elements and animations and all one object centered seats.
And this is where things start to come together. You now have a reference folder with your style and a Figma board that essentially breaks down the editing style into every little element that makes it up. So now we head into Cloud Code and start by uploading our reference folder with our 10 edits.
Then we upload our Figma board with the editing style breakdown. And then this is the prompt that we write to get it to study the full style. I want you to learn this premium minimal editing style.
We've uploaded 10 references of what this editing style looks like and a full Figma board breaking down its design fundamentals. First, take the references and do a frame by frame breakdown with word -by -word transcript on each of the 10 videos.
Then, study three things. First, analyze the designs, colors, layouts, effects, proportions, and fonts. Then, analyze the animation logic, types of animations, timing, speed, and motion curve.
And finally, analyze the animation behavior in relation to the script. After you do all of that, I want you to scan the Figma board we uploaded and use the full breakdown we've given you of every scene to write a complete rulebook of this editing style. And after doing all of this, create the minimal short -form editing skill, the skill that can take raw videos, and with our help, in planning the edit to create the stunning edits in this exact editing style.
Then I just click on enter and wait to see the result. Now here's what's happening in the background. Quad takes all 10 videos, breaks them down frame by frame, and transcribes each scene script.
Then he looks at each animation, studies the design, pulls the color palette, the fonts, the shapes, the proportions, and everything. And he does this for every single animation in the video until he ends up with a folder with all 235 animations broken down step by step.
Then he takes our Figma board and scans through every element on it. He learns the breakdowns and puts together a full rule book of what this style is made up of and how every scene is designed. And using the breakdowns he created and the knowledge he learned from the Figma file, he puts together a skill file that knows everything there is to this editing style.
But what's this skill file made of? How does it know how all these designs are made and how all these animations behave in one file.
If you double click on it, the first thing you're going to find is your breakdown file. This contains the raw, deep frame by frame study of each of the reference edits sent in. Then you're going to find the inspiration guide.
This folder contains 14 categories of scene types, each containing every animation in the scene type found in the video and every animation broken down into a design still, a keyframe sheet and a breakdown file of how it's made. This is done for a total of 235 animations that make up the style. Under that file, we find the lookbook file, which contains the breakdown of every color palette and font style for this style.
and how the skill is meant to use it. Then under all of that, we have the skill file, the actual file responsible for taking all of this information and using it to create a stunning edit in this style. So not only do you have a reference of every single animation in this editing style, deep studies on the typography, color theory, and animation theory of the style, but you also have a skill that knows how to use all of that information to create stunning edits in this style.
And this is how you build a skill that creates stunning short form edits that look amazing. Now I know what you're thinking. This sounds like such a complex process.
Do I need to - do this every single time I want to train Claude on a skill? You actually don't need to worry about any of that.
Since I spent so long training Claude on new editing style for these videos, we package each one into a full skill pack that you can upload to your own Claude. Now, for the skills that you're seeing right now, we're reserving the skill for our AI editor program that's releasing soon. But inside Ultimate Editors, I've already released five stunning editing style skills that our editors have been using to create stunning edits.
The Emangaji editing skill pack, the Vox documentary skill pack, the short form viral editor skill pack, and the two skills that edit my own short -form content and long -form content YouTube videos. Those last two come with the Joseph AI Editing Masterclass, where you're gonna get step -by -step courses on how I train these skills and how I use them to create edits for my short -form and long -form content.
Again, there's just a few more days until it closes down forever. So click the link down below and get the AI skills while you still can. Now that we've created the short -form editor skill pack, it's finally time to put it to the test by taking a raw video and editing it in this stunning style entirely inside of Clot.
Let's go ahead and do that. Now, this is the raw clip. I'm going to be editing.
It's clipped out of a podcast with Chris Williamson and Andrew Huberman, and it starts off with Chris asking a question and then Andrew Huberman answering it. This is the same format as our references, so let's go ahead and edit it. First, I load up Claude and set the model to the new Fable 5 .2 and set the effort to high for this edit.
Then I upload the raw clip into Claude. Now, this clip is horizontal and has all kinds of silences that need to be removed. So let's start by turning our raw video into a clip that can be edited.
First, I ask Claude to do a word by word millisecond level transcript. Then create a frame -by -frame breakdown of the video. Cut out all the silences from the clip and bring it down to 60 seconds long.
Then turn it into a vertical 9x16 video for Instagram, always keeping the speaker in the center of the screen. To ensure that no one gets out of the screen, simply track the nose of the speaker and always keep it aligned with the vertical center of the screen by tracking the speaker. Then deliver the full raw clip ready to be edited.
After that, Claude goes and transcribes the entire video and creates a word -by -word transcript and frame -by -frame breakdown. Using the transcript, it analyzes where the silences are and anything that needs to be removed. cuts it out and turns it into a raw clip.
Then it looks at the frame by frame breakdown, crops it so that the subject is always in the center of the screen and tracks him through every single cut. After that, it'll send me back the full raw cut up video that should look like this. It's nine by 16.
The subject is always tracked and the cuts are clear. Now we can move into the most important part of the video, the planning. Now I'll prompt Claude asking it to use the transcript it just created to create a planning document for me.
On the right side, I want to see each line of the script and on the left side, I want empty boxes where I can plan out what my scene needs to look like. Then I just click on it. and one thing that i learned by editing with ai is humans are still better at planning how the edit should go and what the viewer needs to see on the screen and i always get better results that way so how do i do my plan take a look at this the first part of the planning document asks me what color palette i want to use and font family i chose to go with a golden orange color palette and a sans font family these look the best now the first clip we get started with has this what's your best advice for improving memory and focus chris starts by asking the question in this style every time a speaker starts we write out the words in the hook and introduce the speaker in the lower thirds.
So I plan to write the captions on the screen, emphasize the words improving memory and focus, and have a clean white character introduction in the bottom left. Our next clip says this. specific thing it's an inability to focus again when a new speaker is on the screen captions in the middle and speaker introduction in the lower third then for the first scene andrew says this people with severe clinical grade adhd can focus if they're doing something they really like i chose to hook our viewer in by popping up three different images of a character focusing working and doing something he loves each one using a selector box to highlight the character with an arrow pointing to it and text describing the scene an animation early like this in a video is going to make the viewer really think that this video is about to be amazing to watch.
Then Andrew Huberman says this, They're functional, but the threshold to access them is too high. So I chose to go with a full screen animation to visualize this.
Three circles pop up. Each one contains the molecule, cast a shadow on each circle and design elements on the edges. And then Andrew Huberman says this.
What does that mean? What it means is that people can learn to focus. This is a transitional statement, and it's a good idea to keep these as a roll with subtitles to help transition the viewer into the next scene.
Now, I'm not going to bore you by showing you how I planned every single scene, but here's a macro view of what we did. When Andrew Huberman breaks down the components of focus, we chose to go with. b -roll scene and design elements around the character with text another transitional statement so we went back to a -roll when we talk about sleep we create a sleeping in the cloud scene like this and when we talk about exercise and exercise bureau with text and graphics when andrew huberman talks about how it increases the molecules we pop them up on the screen in a pattern showcase and when he breaks down the type of exercise we go back to b -roll and text again then finally when he says the conclusion we showcase the same molecule animation and threshold but the molecules slide up and break through the shutter and finally we end the video on a -roll as fay one thing to note is in order to plan a good video you need to have some solid experience in editing good videos so that you can understand what the viewer should see and feel based on each section this is a very special skill that you need to build especially with ai rising so fast now that the full plan is built we can copy it and drop it on cloud and tell it to edit the video and this is the best all of our manual work is done so we could just sit back and just watch cloud do all the work first cloud starts by collecting the assets it looks through all the planned out scenes and puts together a list of assets it needs to collect then using the chrome extension it goes into chrome and downloads anything it needs
images, illustrations, B -roll, and icons. After it has collected the asset list, it will then start by building out our captions and subtitles. It uses our font studies and creates the captions for the entire video and then sends us this.
This design sheet shows what our captions look like and gives us the opportunity to tell Cloud to fix anything that went wrong. And check out these caption designs. Just look at how good they all look.
After that, it starts working on the B -roll segment and creates the designs with the selectors. Just look at how high quality these all look. And then it builds out the design for these molecule pop -up scenes using the foundation.
studies it did on our style earlier and this looks perfect then it builds the two focus scene with the graphics and sends them in they both look perfect and are perfectly nailed and after that it goes and builds the sleep scene which looks solid then it builds out the exercise scene that also looks perfect after that it builds out the molecule showcase scene which also looks perfect and finally it creates the breakthrough molecule scene for the final segment which looks awesome after wrapping up all the animations in about 45 minutes it puts them together into the video and begins what's called the final render now the final render just exports all the animations into the file 1080p video and syncs them up with the voices too and finally it sends us this the final export this is our opportunity to watch through it and confirm that it looks good before we proceed so i watched through it and it looked perfect which now meant there was one last thing to do and that's the sound design and music now i've already trained claude on the sound design workflow for this style so i just ask it to begin the sound design process and that's when it starts by doing an audit for the full video it breaks down what happens on the screen and at what time and what corresponding sound effects should be used for it after it does that it adds the sound
effects into the video at the right time and volume accordingly. After doing that, it'll pick out 10 soundtracks that fit the video and send me 10 second previews of what each one would sound like. I listen through the 10 soundtracks and I choose the best fitting one.
It adds it in, renders the 1080p master, and this is the final result that you end up with after training Cloud on the minimal podcast editing style. What's your best advice for improving memory and focus? Brain fog is not a...
Specific thing it's an inability to focus people with severe clinical grade adhd Can focus if they're doing something they really like the dopamine norepinephrine epinephrine are there They're functional, but the threshold to access them is too high. What does that mean?
What it means is that people can learn to focus and they can get assistance in doing that by pushing on one of two ends or both. If somebody wants to learn, you want to focus and get rid of brain fog, here are the key steps. You need to focus, obviously, to learn.
To focus, you need to be alert. To get alert, there are many things you can do. There are foundational things like sleep that's going to help you get alert and focused, but there are also things like exercise.
Exercise makes better at learning for two reasons one it increases the catecholamines dopamine epinephrine and norepinephrine 45 to 75 minutes of hard training not taken to extreme you're going to come out of that more alert because you have more adrenaline in your body more norepinephrine being released in your brain and more dopamine being released in your brain and dopamine is a critical component for learning okay if there's one big takeaway from this video it's this as a video editor you need to understand how good content is edited in 2026 and you need to learn how you could do that using ai because in 2020 the editors who learn AI video editing are going to be so far ahead that for the editors who didn't, it's going to be way too late for you to catch up.
This is why with only a few days left, this is the best time to join Ultimate Editors 2 .0. Inside, you get five step -by -step masterclasses on the viral editing styles of 2026, the cinematic style, the minimal animation style, the SAS trailer masterclass, the viral animation breakdowns, and the viral talking and masterclass.
The AI editing fundamentals course will teach you everything you need to know about AI as a video editor from A to Z. And then the Joseph AI editing masterclass will give you access to the skills I use to create viral short form edits on my Instagram and the very YouTube videos you're watching right now.
On top of this, we offer a $2 ,000 per month editor blueprint where we teach you the secret techniques that editors have been using to make $2 ,000 per month with video editing in 2026. And this is how we've helped editors make a total of $282 ,000 over the last nine months. Look, all of this closes down forever on November 15th, and you can now get lifetime access for just $299.
This includes all the AI modules and all the skills that you need. This is officially your last chance, so take it while you can.
Join Ultimate Editors 2 .0, the link is down below. I'll see you inside.
The Hook
The bait, then the rug-pull.
Joseph shows how he used Fable 5.2 to edit a premium Instagram Reel with nothing but a Claude subscription, no Higgsfield and no deep editing knowledge. The video moves from the four design principles, to training Claude on a style, to editing a real clip, then closes on a course pitch with a hard November 15 deadline.
Named ideas worth stealing.
00:58list
Four principles of premium talking head edits
- Colors and fonts
- A-roll and subtitles
- B-roll scenes
- Full-screen animations
The four building blocks that make up the premium minimal talking head style.
Steal forteaching an AI or a new editor what to look for in a style
01:27model
Four-color palette from one accent
- Bright accent (gold)
- Dark accent (orange)
- Tinted white
- Tinted black
Pick an accent color from the footage, extract a bright and a dark version, then tint white and black toward it.
Steal forany branded video or slide template
04:52list
Four kinds of full-screen animation
- Two-object alignment
- Graphs and numbers
- Real elements in animations
- One object centered
Scene setups chosen by what the speaker is talking about at that moment.
Steal formatching visuals to script content
08:18list
Train-Claude-on-a-style workflow
- Collect 10 references
- Build a Figma inspiration board
- Break the style into sections
- Have Claude study both and write a skill
The repeatable method for teaching Claude any editing style.
Steal forany design or content style you want an AI to reproduce
How they asked for the click.
VERBAL ASK
21:02product
“All of this closes down forever on November 15th, and you can now get lifetime access for just $299.”
The pitch appears three times (around 0:30, 6:50 and 13:24) before the full close, each time with the November 15 deadline, so the scarcity is repeated throughout the tutorial.
Visual structure at a glance.
