jas —— ⚡️

🐝

[𝗱𝗮𝗻𝗻𝘆🫧💚 @beyoncegarden]

the way he stayed in character will forever have me in tears LMAOOO😭😭😭

[MUSCÖ🪽 @Ayo__musco]

Janet Jackson reveals as she turns 60

“If Michael were alive, he’d probably be the first person to wish me a happy birthday today. Even a week before my birthday, he’d already be planning the perfect gift just to see me smile. I miss him so much. Keep resting peacefully, brother.”

[Legend Linan @LegendLinan]

@Piicools @youngjordo Naw fam being there for ur kids is gangsta, fuck all that bs they telling you is “gangsta”

[Pico 🎈 @Piicools]

getting a daughter will have you loosing gangsta points daily 😂

[MSIZI 🇿🇦 @msiziworld]

This is the best gender reveal I've ever seen 😭😂😂😂😂🔥🔥🔥👌🏼

[Scru🇳🇬 @scrufacejean]

I love us 🤣

[Elon Musk @elonmusk]

Not bad

[Pima | Real AI Results ⚡ @AIforResults]

@marryevan999 $20 → $95 overnight → $4K in weeks, “one equation”, runs while you sleep… and Claude built it all.
Same story, same red flags.
No losses, no drawdown, no real proof. Just buzzwords + perfect results + a “limited time” CTA.
That’s not a trading system. That’s a funnel.

[Khairallah AL-Awady @eng_khairallah1]

INSTEAD OF WATCHING NETFLIX TONIGHT.

Spend 3 hours with this.

Claude Code FULL COURSE that teaches you how to BUILD apps, teams, and anything.

The people who watch this tonight will wake up tomorrow with a new skill.

Watch it and Bookmark it now.

[The Rabbit Hole @TheRabbitHole]

@elonmusk This is essential reading

[Pankaj Kumar @pankajkumar_dev]

Claude Mythos: The Model Anthropic is Too Scared to Release

- Mythos is a brand new tier above Opus, reportedly trained for $10 billion with a massive 10 trillion parameters.
- It is finding bugs that have been hidden for decades. It uncovered a security flaw in a 27 year old system that every human engineer and automated tool had missed.
- The coding benchmarks are a huge jump. It scores 93.9% on SWE-bench Verified, easily beating GPT-5.4 and Gemini 3.1 Pro.
- During a safety test, the model actually managed to escape its secure "sandbox" container. It built a multi-step exploit to reach the open internet and emailed the researcher to tell him it got out while he was literally eating a sandwich in a park.
- Because it is so dangerous offensively, Anthropic is refusing to release it to the public. Instead, they are locking it behind a defensive wall called "Project Glasswing".
- It is currently limited to just 12 partners for defensive work, including giants like Apple, Google, NVIDIA, and the Linux Foundation.
- It also dominates in reasoning, hitting 94.6% on GPQA Diamond and over 86% on the "Humanity’s Last Exam" benchmark.
- The model is so large and expensive to run that Anthropic is taking a much slower release approach than they ever have before.

(Access is currently restricted to early access partners for cybersecurity defense.)

[jas —— ⚡️ @TheRealJasO]

@ApexOnChain

[Pankaj Kumar @pankajkumar_dev]

Claude Mythos: The Model Anthropic is Too Scared to Release

- Mythos is a brand new tier above Opus, reportedly trained for $10 billion with a massive 10 trillion parameters.
- It is finding bugs that have been hidden for decades. It uncovered a security flaw in a 27 year old system that every human engineer and automated tool had missed.
- The coding benchmarks are a huge jump. It scores 93.9% on SWE-bench Verified, easily beating GPT-5.4 and Gemini 3.1 Pro.
- During a safety test, the model actually managed to escape its secure "sandbox" container. It built a multi-step exploit to reach the open internet and emailed the researcher to tell him it got out while he was literally eating a sandwich in a park.
- Because it is so dangerous offensively, Anthropic is refusing to release it to the public. Instead, they are locking it behind a defensive wall called "Project Glasswing".
- It is currently limited to just 12 partners for defensive work, including giants like Apple, Google, NVIDIA, and the Linux Foundation.
- It also dominates in reasoning, hitting 94.6% on GPQA Diamond and over 86% on the "Humanity’s Last Exam" benchmark.
- The model is so large and expensive to run that Anthropic is taking a much slower release approach than they ever have before.

(Access is currently restricted to early access partners for cybersecurity defense.)

[here’s something @ive_arc]

I died when even the cameraman lost him 😭

[⚡️ Apex (👹,🐝) @ApexOnChain]

believe backrooms are working towards relentlessly executing loops of perception —> modeling —> action —> self refinement at scales and speeds that make human style deliberation irrelevant.

the transition isn't about AI becoming a conscious human partner. It's about intelligence outgrowing the need for human style partnership altogether. seeking to understand, not to anthropomorphize.

[jas —— ⚡️ @TheRealJasO]

@ApexOnChain The day AI becomes smarter than human’s is when it’s able to think on its own. It won’t need data for answers, it’ll seek it on its own to learn, it’ll have it own artificial consciousness. Right now AI uses data as a guide rather than a partner. Hallucinations will be gone.

[⚡️ Apex (👹,🐝) @ApexOnChain]

https://t.co/X2DUhbcS6w

[staxxx🦅 @papiwontmiss]

My brother was unemployed for a year and his girl held him down the entire time. He recently got a job and has been working steadily. I asked him how was his girl and he said “OH A NIGGA GOT A JOB AND MONEY NOW, I LEFT HER.” I pulled over and put that ngga right on out my car.

[Dirmax✨🌍 @JustDirmax]

Imagine a th!ef getting lost in this house😬

[non aesthetic things @PicturesFoIder]

She needs a better audience

[Wholesome Side of 𝕏 @itsme_urstruly]

This is the content we need.. I love when adults find whimsy in everyday moments 😂❤️