โœจ Visual Editor

close

palette Canvas & Background

Gradient:arrow_forward
Text Color:
135ยฐ

style Card Style

40px
16px

text_fields Typography

16px
phed
@PhedEU
This channel started 3 months ago and is already making ~ $15,000/month

Here's how I would enter this niche with AI:

A thread ๐Ÿงต๐Ÿ‘‡
Thread image
phed
@PhedEU
Before we get into it, I post a lot of YouTube related threads

Bookmark to come back and re-read this thread in the future ๐Ÿ‘‡
phed
@PhedEU
First Stage: IDEATION

- We will use ChatGPT o4/o4-mini for this

- Your first prompt will be:

"You are an elite youtube strategist that handles ideation for faceless youtube channels.

Attached is a viral channel's videos in the philosophy niche. Give me video ideas for a channel I'm launching in the same niche"

And attach an image of the most viewed videos of the channel
Thread image
phed
@PhedEU
You will get your first couple of ideas

You can ask AI to only use 2-3 "famous" philosophers like Plato, Seneca etc
phed
@PhedEU
Second Stage: SCRIPT WRITING

We will use the same GPT model (o4)

You can stay in the same window you just used to ideate

Prompt will be as follows:

"Below is the script of a video called "Nietzsche EXPOSED Truth about some Women And No One Listened!" from this channel:

* PASTE SCRIPT HERE *

Here is a script of a video called "Carl Jungโ€™s SHOCKING DARK Truth About SOME Women Men Must Know!"

* PASTE SCRIPT HERE *

Now using this writing style/pace/engagement tactics, write a script for a video called "USE ONE OF THE IDEAS IT GAVE YOU ON STAGE ONE""
Thread image
phed
@PhedEU
Third Stage: Image Generation

- This channel is using "oil painting" images

- You can get sentences out of your script to generate the image. A simple prompt would be:

"Generate an image of this concept "QUOTE FROM THE SCRIPT" in a hand-painted oil-painting style, aspect ratio 16:9"
Thread image
phed
@PhedEU
You can generate 2-3 images per paragraph

If you wanna go the extra mile, animate them using RunwayML/any other Image to Video model
phed
@PhedEU
Fourth Stage: THUMBNAILS

This channel is using the "Surrealism" style for its thumbnails

Here's how a prompt would look like:

"generate this image: A haunting surreal landscape at dusk, painted in muted yellows and browns like aged oil on canvas. In the foreground, a lone small figure stands before a reclining woman in white undergarments, her legs raised skyward. Beyond her is a looming mountain made of countless identical bald male heads, each detailed and expressionless. Moody, painterly brushstrokes, high texture, dramatic low light, cinematic composition.

aspect ratio 16:9"
Thread image
phed
@PhedEU
This was my attempt in remaking his most popular video's thumbnail
Thread image
Thread image
phed
@PhedEU
A good practice to learn how to generate such thumbnails is to ask ChatGPT how it would prompt attached picture - example attached:

You will very soon learn how to prompt better and better
Thread image
phed
@PhedEU
You can create SOPs around all these methods and hire VAs to run these channels for you
phed
@PhedEU
If you are still reading, drop a follow as this took a while to write

And as always, join the newsletter at ytdojo.io
Generated by Thread Navigator
100%
view_carousel Carousel Studio NEW
Press โŒ˜ + S to quick-export