#ONEPIECE1189
" القرصان صاحب رقعة العين الذي ذكره أودا هو لوفي ! "
أودن في الأساطير النوردية طُعن على شجرة تماماً كما حدث مع لوفي ونال لقاء ذلك المعرفة المطلقة والقوة وأسرار الكون
و ايضا ضحّى بعينه اليسرى في " نبع ميمير " مقابل الحصول على الحكمة والبصيرة
وبما أن أودا جسّد أسطورة أودن مع لوفي من خلال حادثة طعن الشجرة، فمن الواضح أن الجزء الآخر من الأسطورة ( فقدان العين اليسرى ) سيحدث مع لوفي أيضاً
لكن كيف سيفقد لوفي عينه ؟
- في نهاية الفصل رأينا لوفي يسقط نحو العالم السفلي في إلباف، وقد يفقد عينه نتيجة هذا السقوط ، أو بسبب تأثير أحد الأسلحة المغروسة في جسده ( سأُفصّل فيها في تغريدة منفصلة )
- قد يفقد لوفي عينه بسبب اوسوب كما رسم اودا في عام 2019
- قد يكون إيمو هو السبب المباشر في فقدان لوفي لعينه
Art by : @iamosix2
One Piece fans came up with a theory that Imu is a parrot and it’s actually pretty funny when you think about it So Roger laughed at Laugh Tale because he found out the ruler of the world was actually a parrot then Roger said “It’s such a funny story” fck lol😭😭
New in Claude Cowork: teach Claude a skill.
Record your screen while you do a task, talk through it as you go, and Claude turns it into a skill it can run again. Find it under Record a skill in the + menu of the Claude desktop app.
Available on Pro, Max, and Team plans.
Kimi CEO Zhilin Yang:
"Claude thinks more data is the edge - it's not - the data's already gone
the real edge is wasting less of it - 2x the intelligence from every token - that's Kimi 3"
his one big idea: token efficiency is the new Moore's Law - twice the intelligence, same data, no extra GPU
- he threw out Adam, the optimizer every lab's run since 2014. 15 trillion tokens, not one spike
- a curve he called "the most beautiful thing I've seen in 2025"
the real goal? "every token a model generates is non-fungible" - a CEO's mind isn't a designer's. taste is the frontier nobody's building for
watch & bookmark
Instead of watching 2 hours of Netflix tonight, watch this in-depth interview with Kimi founder Zhilin Yang.
He explains why everyone’s racing on reasoning while Claude quietly bet on agents - but the real hard part isn’t the agent itself, it’s the model underneath.
His memorable metaphor: a pure reasoning model is like “a brain in a fish tank” - it can think forever but never touches the world. An agent is that brain wired to tools, memory, and action.
Yang shares how they’re thinking about K2 helping build K3 through strong agent skills, the biggest bottlenecks in generalization, test-time scaling in two dimensions, and even how RL-style thinking applies to both models and running teams.
Whether you’re building agents, scaling models, or just want to understand frontier AI thinking beyond the usual Western narratives, this is one of the clearest and most thoughtful long-form discussions available right now.
Bookmark it & watch - it might be the most valuable 100 minutes you spend on real AI engineering and strategy this month.
Kimi's CEO Zhilin Yang:
"OpenAI, Google, everyone just kept scaling data. that
game is over - we ran out of internet."
here's the shift he's betting on:
→ scaling data hit a wall - the high-quality internet is basically used up
→ O1 opened a new axis: scale the thinking at inference, not just the training data
→ his bet: the training-to-reasoning ratio flips, and that's where the next startups win
bookmark this ↓
BREAKING: 🇺🇸🇨🇳 China is now leading the AI race after spending ten times less of what the US did
Kimi-K3 by @Kimi_Moonshot is now number 1 in the Frontend Code Arena with 1679 pts, surpassing Claude Fable 5.
It's honestly disgusting how openly retail gets farmed here.
I'm not bearish these companies. I just think the valuations US retail gets handed on day one are ridiculous.
All the reasonable-entry upside already happened in private, walled off from us. By the time we're allowed in, the tag reads:
- SpaceX, ~$1.8T at IPO
- Anthropic, ~$1T, still private
- OpenAI, ~$850B, still private
- Cerebras, ~$80B on day one
Now look at where Chinese retail gets to step in:
- Zhipu, ~$7B at IPO
- MiniMax, ~$6B at IPO
- CXMT, ~$80B with the move still ahead, and the public can actually subscribe
American retail only gets access once the valuation is already in the stratosphere.
The 100x always happens where you're not allowed to be. You only get invited when it's time to hold the bag.