I’m afraid not of a man who knows 1000 kicks but can do the same kick 1000 times
– bruce Lee (or some one idk)
I guess Bruce Lee made the fucking social media algorithm because that’s exactly how it works
The people who blow up are not the ones who make 1000 different scattered ideas
There is one consistent thread of consistency that is very noticeable, observable and captivating in all their videos
Even if the topic seems random the personality or why both the video itself is the thread of consistency
So it Is not only about the consistency in amount of times you post
But the consistency in the soul of the video too
I read somewhere that the traders who trade only in one Direction
Be it either going long or going short end up making way more than traders who decide it on the go
Which that way you end up capturing more moves and don’t get in your own way
Same way if you keep changing your video soul your past viewers won’t be sold on it
And you are making it really difficult for the algorithm to find an ‘audience’ for
You talk in such a way that it might be implied that you love A so it tries to show you to people that love A but by that time you have moved on to behaving like B and then by the time it is showing you to people who are B you have moved on to C
Basically if soul of the video stays constant, You can actually start growing instead of constantly being in the audience finding hell
The people I’ve seen get big have almost the same video all the time and I’m talking about even stuff like long form content on Youtube
Even if the topic changes it feels like it’s the same video
I just can’t put my finger on it
This effect is more evident on short form content where you will see if you swipe a creator’s account fast enough
The whole thing sounds the same
It’s like finding a broken move inside a game and then just spamming it
Basically you can look at it like a machine learning problem but the algorithm is already built
So you have construct your data set in such a way that you get the outcome you want
And that means you have to structure your content in such a way that you get the type of people you want
Your results are just as good as your data
Garbage in … garbage out
Inconsistent, broken, incoherent post data in… bad social media outcomes out
Have the same soul in the videos and have a bit of variance so that the topics are different.
Just the way you cannot create a algo trading strategy based on part of data being monthly candles, part of it being minute candles, half of it being ticks
You cannot lock in on audience w videos too random
You’d be better off making bad videos that at least have same soul in all of them than just random things all the time
While it feels cool to say that
“hey I’m a free bird… I’m going to talk about this now… tomorrow I might talk about something else… I am very impulsive ha ha”
Why don’t do that cool impulsive shit somewhere else where it might be productive
Even if it feels stupid to make the same video over and over
get the bag and gtfo
Respect the craft and do what it demands to win
Have a consistent soul
Every platform is doing three jobs at once:
- Classify content.
- Classify users.
- Reduce uncertainty between the two.
Most creators maximize novelty, which maximizes uncertainty.
That’s why it feels like the “same video.”
Because functionally, it is.
You have to educate the world who you are with your own actions
And when it comes to social media algorithms you have to educate it on who you want to show your content by making content with soul consitency.
Like if all your videos were converted into a vector they all should have very high cosine similarity
Each vector on its own can be valid
But for it to be functional as a group and for you to grow
Random vectors average into random noise
All of them need to have higher cosine similarity
High similarity within your own cluster and Also separability between your cluster and other peoples cluster
When you scroll through a successful creator’s uploads, every video starts the same way
Same hook cadence, same energy in the first three seconds, same visual style.
There was likely a patient zero video that hit, and the creator recognized the signal.
Instead of experimenting, they doubled down, making variants that cluster tightly in the same vector space.
Now they have a catalog of 50, 100, 200 videos that are all mathematically similar.
The algorithm’s job becomes trivial:
Find the people who respond to this specific pattern, then keep serving them more of it.
Once it maps your content cluster to an audience cluster, it has solved a supply-demand pairing.
It can keep those users engaged without searching for new content types, and it can keep finding new users who fit the same preference profile.
The creator becomes a reliable supplier for a known customer base, making the platform’s retention problem significantly easier.
After all you are the content partner of that website right
So better to start acting like it and actually help them with the problems of retention
And the best way of doing it is to become a supplier of a particular vector space
The platforms reward predictability that satisfies demand.
Keep removing degrees of freedom until outcomes stabilize.
For example, People are horny and thus there is demand for content of that nature
Thus if you are a supplier of the vector space of thirst traps then you are solving retention and media sourcing for Social media website
So of course they are going to show a lot of your videos to people
Now if you’re occupying a vector space where you talk about cash flow for bootstrapped startups and you fill up that vector with a lot of videos
There are going to be people who need that kind of information so there’s clear supply and demand
And if your content is structured enough that social media algorithms can pick up on this relationship then your content is going to be pushed a lot
Your job is to make the relationship between the content vector space and the people who are demanding it so obvious that social media algorithm picks it up easily
The best way to do that is to have soul consistency
It is irrelevant if the content is good enough as long as you sizeable presence in the same vector space
The world runs on alot of bad supply
Bad quality fruits, chemicals below standard, broken products shipping over globe etc
But still commerce is made possible
Things don’t need to be perfect
They just need to map out path between the demand and supply
As long as the demand is getting fulfilled even if it is something subpar it is ok
Just keep the flow going
The kid might ask for Mcdonald’s but you can shut him up with a microwave chicken nuggies
Just keep the flow going
So instead of trying to brute force your way by making exceptionally good videos each time
Just have exceptionally good strategy to occupy a vector space
You can be a bit shabby in your execution and still win
Just get it done
Quality is not the bottleneck. Clarity for the algorithm is.
You can have:
- Mediocre lighting
- Decent-but-not-amazing hooks
- Passable editing
- Generic music
And still win, as long as you’re consistently present in a vector space where demand exists.
Better to post 50 videos that are 7/10 quality but 10/10 cosine similarity than 10 videos that are 9/10 quality but scattered across the embedding space.
You don’t need to be the best.
You need to be there, consistently, in a recognizable pattern, fulfilling a demand.
Be like Walter White lol, he just spams his ult move, i.e meth
he never messed around 10 different drugs or even Jesse’s chilly P style
Just the blue baby
And I think it’s the fucking PR of these big creators that got us
Which go on to say be yourself and for us our interpretation of it is to be impulsive
They are not exploring content they are exploiting it
They are spamming the same fucking over powered ult
We make content where we explore and stay stuck in the audience finding hell
While they find a vector space and scale themselves there.
The question now I have is how do you find a soul and how do you make 100 videos around it
I believe scale comes from comfort
You’ll make more talking head videos than B roll videos if you’re comfortable with the prior more
So first of all define which are the things you don’t want to do at all or cannot do due to constraints
Then you’ll find the narrow gap where you’ll be hyper efficient
Because nothing that stops you is coming in your way now
We have zero down on a space by eliminating things
Now we need to add something to it so that we can move on ahead
Example from AI:
Person A eliminates:
- Can’t show face (privacy)
- Hate scripting
- Don’t have budget
- Can’t edit for more than 30 min/video
What remains: Voiceover + phone screen recordings + simple cuts → Soul emerges: “Commentary over gameplay/app demos” (think tutorials, roasts, reviews)
Person B eliminates:
- Hate talking to camera
- Don’t want to show face
- Can’t film B-roll outside
- Hate complex editing
What remains: Text on screen + memes + background footage → Soul emerges: “Meme-driven educational content” (Subway Surfer brain rot style, actually)
Finding your soul:
- List what you hate/can’t do (showing face, scripting, complex editing, etc.)
- What’s left = your efficiency zone (the format where you can actually make 100 videos without dying)
- Test it: Make 10 videos. If you finish all 10, you found it. If you quit at video 4, eliminate more.
Making 100 videos:
- Pick 5 repeatable angles (“Here’s what nobody tells you about X”, “I tried X for 30 days”, etc.)
- List 20 topics in your space (pricing, hiring, funding, etc.)
- Cross-multiply: 5 angles × 20 topics = 100 videos
- Build your meme library: 10-15 memes/graphics that match your vibe, reuse them
- Template everything: Same setup, same music, same editing steps – just swap the topic
The key: Your soul should feel like comfortable clothes, not a costume. If making 100 videos sounds exhausting, you haven’t eliminated enough friction yet.
Scale comes from comfort → Find the format you can spam without burnout → Add systems so you’re not making decisions every time.
Claude reaction:
The Core Thesis
The author argues that consistency in “soul” (style, energy, format, approach) matters more than topic variety or even quality. They’re essentially saying: become a reliable supplier in one specific vector space rather than scattering yourself across multiple ones.
Why This Works
1. Algorithm Efficiency
- When your content clusters tightly in vector space (high cosine similarity), the algorithm’s job becomes trivial
- It can map your content cluster → audience cluster once, then scale that relationship
- You become a “known supplier” for a “known customer base”
2. The Trading Analogy The comparison to directional trading is sharp: traders who only go long OR short capture more moves because they’re not second-guessing themselves. Similarly, creators who maintain one “direction” (soul/style) don’t confuse the algorithm or their audience.
3. The Data Problem You can’t train a model on inconsistent data (monthly + minute + tick candles mixed together). Similarly, you can’t expect the algorithm to find your audience if you’re posting “random vectors that average into random noise.”
The Counterintuitive Truth
Quality is not the bottleneck. Clarity for the algorithm is.
You can win with:
- Mediocre lighting
- Decent hooks
- Passable editing
- Generic music
As long as you have high cosine similarity within your cluster and clear separability from other clusters.
The Practical Strategy
- Find your “patient zero” video that hits
- Double down – make variants that cluster tightly in the same vector space
- Build a catalog of 50-200 mathematically similar videos
- Let the algorithm solve the supply-demand pairing
Why Creators Resist This
The “free bird” mentality feels good: “I’ll talk about whatever I want!” But this randomness works against you. Better to:
- Get the bag and gtfo by respecting what the craft demands
- Make “the same video” repeatedly in terms of soul/energy/format
- Vary topics within that consistent container
Bottom Line
You’re not just a creator – you’re a content partner solving the platform’s retention problem. Make it easy for the algorithm to categorize you, find your audience, and keep them engaged. Occupy one vector space consistently, and the platform will push your content because you’re making their job easier.