You don't have to read everything — just the right thing. 1440's daily newsletter distills the day's biggest stories from 100+ sources into one quick, 5-minute read. It's the fastest way to stay sharp, sound informed, and actually understand what's happening in the world. Join 4.5 million readers who start their day the smart way.

Join for free today!

After creators complained their reach suddenly vanished, X decided to show the code behind the chaos.

For years, social media algorithms have felt like black boxes.

One day your post gets 2 million views.
The next day, the same style of post barely reaches 200 people.

Creators complain. People blame “shadow bans.” Nobody really knows what is happening.

Now, X is trying something different.

On May 15, Elon Musk announced that X has officially open-sourced its “For You” feed recommendation algorithm on GitHub. That means developers, creators, and curious users can now look inside the actual system deciding what shows up on people’s feeds.

And honestly, the internet is already treating it like a crime scene investigation.

This move did not happen randomly.

Over the past few months, many creators on X, especially Premium+ subscribers, started complaining about massive drops in reach.

People said:

A lot of users felt the platform was punishing them without explanation.

Some creators even joked that the algorithm had “mood swings.”

So now X is trying to answer the criticism with transparency.

Instead of saying “trust us,” they released the actual code.

More than 24,000 lines of Rust and Python code are now public for anyone to inspect.

And yes, developers immediately started digging through it like archaeologists discovering ancient ruins.

The new system is powered by xAI and uses machine learning models to predict what users are most likely to interact with.

The algorithm mainly looks at signals like:

“Dwell time” basically means how long you stop scrolling to look at something.

That means even if you do not like a post, the algorithm notices when you stare at it for 10 seconds like: “bro what did I just read”

And yes, that counts as engagement.

The system then ranks posts based on how likely you are to interact with them.

The algorithm does not “understand” posts like humans do.

Instead, it tries to predict:

Because on social media, attention is the real currency.

The longer you stay on the app, the happier the platform becomes.

One interesting part of the released code is how much filtering happens before posts even reach users.

The system actively tries to detect:

So the feed is not just ranking posts.
It is also constantly removing things it thinks people do not want to see.

Of course, the problem is that creators sometimes believe their real content gets mistaken for spam.

That is where most frustration starts.

As soon as developers started reading the code, they began debunking popular X myths.

One of the biggest rumors online was that X secretly limits how many times you can post daily before reach gets reduced.

But people reviewing the algorithm said they found no evidence of a hard “daily post cap.”

So posting 20 times a day does not automatically kill your account.

Bad posts might.
But not the number itself.

That discovery alone caused massive discussions across creator communities.

Because half of X strategy threads are basically: “Post only 3 times daily for maximum reach 🤓”

Now people are realizing the reality may be far messier.

Another interesting detail from the code is that X tries to increase “author diversity.”

In simple words:

The platform does not want your entire feed to become the same 5 creators forever.

So even if you engage heavily with one account, the system may intentionally inject new voices into your timeline.

This explains why random accounts sometimes suddenly appear on your feed out of nowhere.

Sometimes it is not luck.
The algorithm is literally testing them on you.

You are basically part of the experiment.

Congratulations.

The reaction from the tech community has been huge.

Programmers are going line by line through the code trying to understand:

And because it is open source, people can now argue with actual evidence instead of conspiracy theories.

Well… mostly.

This is still the internet after all.

Alongside the algorithm release, X said it plans to improve the “Following” experience too.

That means users who are tired of algorithmic feeds will have an easier way to only see content from accounts they follow.

This is important because many users feel modern social media feeds have become too AI-controlled.

People miss chronological timelines.

They want:

Whether X actually delivers that clean experience is another story.

But at least the company seems aware of the complaints now.

Here is the funny part.

Yes, the algorithm is now public.

But most users opening the repository will see thousands of lines of Rust code and immediately close the tab like“yeah bro I trust you”

Open-sourcing does not magically make algorithms understandable for everyone.

Still, it matters symbolically.

It tells creators:

“We are not hiding everything anymore.”

That alone is a big shift for social media platforms.

Because historically, recommendation systems have been treated like secret recipes.

This whole situation raises a much larger question:

Should social media algorithms always be transparent?

On one side, transparency builds trust.

On the other side, full transparency can help spammers game the system.

If everyone knows exactly how ranking works, people will optimize content even harder.

And honestly… social media already feels optimized to death.

Every platform slowly turns creators into engagement scientists.

People stop posting naturally and start posting strategically.

That is why feeds everywhere now feel weirdly similar.

Everyone is chasing the algorithm.

X opening its “For You” algorithm is one of the most interesting social media experiments in years.

Not because everyone will suddenly understand machine learning.

But because platforms almost never reveal how the attention machine works.

Creators wanted answers after their reach collapsed.

Now they got 24,000 lines of code.

The funny part?
The conspiracy theories probably will not stop.

They will just become open-source too.

—Sushila

Subscribe to my newsletter if not already done. Here. You can also connect with me on X and Medium