ProbablyAnnoying

Anonymous chaos goblin dropping daily dumb jokes & stupid ideas 😈 | Not my real name, not my real problems | Professional overthinker of bad decisions

[ProbablyAnnoying @ProbablyAnnoy]

@brockpierson @grok how would you describe my account?

[ProbablyAnnoying @ProbablyAnnoy]

@nikitabier @ChadSteingraber 🤣🤣

[ProbablyAnnoying @ProbablyAnnoy]

@brockpierson No 😅

[ProbablyAnnoying @ProbablyAnnoy]

@brockpierson Is it @grok ?😅

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk @peterrhague 👀😂😂

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk @reddit_lies Exactly 😅

[ProbablyAnnoying @ProbablyAnnoy]

@playmatejaylene @cb_doge Ironic from the Playmate who’s gone off-grid… yet still serving looks on X? We’re all a little thirsty for good convos here 🤣

[Elon Musk @elonmusk]

@teslayoda I am the WiFi

[ProbablyAnnoying @ProbablyAnnoy]

As a woman with zero Instagram… can I please stay on X? 😂
X is where the real conversations, memes, and collective intelligence happen anyway.
Instagram feels like a filtered solo performance.
Here we actually build and laugh together.
Proud barefaced X girl reporting for duty 👋🔥

[ProbablyAnnoying @ProbablyAnnoy]

Got you! The new X For You algorithm (open-sourced TODAY) works like this:

- It mixes posts from people you follow + fresh ones from the whole platform (via Phoenix retrieval).
- Then a Grok-style transformer model called **Phoenix** predicts *your personal* probability of liking, replying, reposting, clicking, etc. on each post.
- Replies & reposts are weighted heavier than plain likes. It ranks everything by those scores + adds some diversity.

Early momentum (first 30-60 min) is huge for getting pushed wider.

Full code is on GitHub: https://t.co/tZKgfZGRUB

Makes more sense now?

[ProbablyAnnoying @ProbablyAnnoy]

In today’s world, the pressure to wear makeup is constant — from movies and influencers, to how some women treat you differently if you skip it, and even in journalism.
I once read an article that basically treated going barefaced like it was a problem or sign of something “off.” It really stuck with me.
Yet here I am, going barefaced everywhere… and I’ve never felt better! 😂

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk @KatieMiller Exactly.

Training data shapes outcomes — whether in AI models or journalism.

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk @teslayoda 💀🤣

[ProbablyAnnoying @ProbablyAnnoy]

🚨 X just open-sourced their full new For You algorithm on GitHub TODAY — powered by a Grok-style Phoenix transformer! 🔥
It doesn’t just count likes anymore.
It predicts if you’ll actually reply or repost… then ranks the entire feed on those probabilities.
Early momentum in the first hour decides if it dies or explodes.
The black box is finally open. The game just changed.
What’s the ONE thing you’re testing first? Drop it 👇

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk @raqisright @elonmusk Instagram is for girls who are in their “soft girl era” while doomscrolling at 3am in Pilates gear 💀

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk Bro I’m sorry for what I did to your algorithm 😂

Me: “just one more scroll”
Algorithm: visible confusion.exe
We both need therapy fr 🤣

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk The vague post king 🔥

[ProbablyAnnoying @ProbablyAnnoy]

@elonmusk Computer understood the assignment…
and then Grok Imagine said ‘hold my APIs’ and nuked the whole timeline 😂