Crypto Musk @cryptmusk

The AI Race Is No Longer US vs China. It’s Closed vs Open I think they’re looking at the wrong thing. This chart shows something much more important: The gap between the leading US frontier models and Chinese AI models is collapsing at an astonishing pace. Just 18 months ago, OpenAI, Anthropic, and Google enjoyed a significant lead. Today, DeepSeek, Qwen, Kimi, MiniMax, Z.AI and others are rapidly closing the gap despite facing GPU and semiconductor restrictions. What can we learn from this? 1️⃣ Constraints drive innovation. History has repeatedly shown that limitations often force teams to become more efficient, creative, and resourceful. 2️⃣ Frontier intelligence is becoming commoditized. The difference between the best and the tenth-best model is shrinking faster than most expected. 3️⃣ The next competitive advantage won’t be the model itself. It will be who can best combine AI with proprietary data, workflows, agents, distribution, and execution. The biggest opportunity over the next 24 months may not be building another model. It may be building businesses, products, and AI operating systems on top of increasingly abundant intelligence. The AI race is no longer just a battle between the US and China. It’s becoming a race to transform intelligence into outcomes. That’s where the real value will be created. What’s your take? Are we entering an era where AI models become commodities, just as cloud infrastructure did?

Christophe Barraud🛢🐳 @C_Barraud

2 hours ago

🇨🇳 🇺🇸 China is rapidly closing the AI gap with the United States - Artificial Analysis artificialanalysis.ai/#frontier-lang…

0 12 20 6K 8

Paying the Bill. Four Countries. Four Mindsets. 🇨🇳 China: Fight to pay first. 🇯🇵 Japan: Split it fairly. 🇰🇷 Korea: The senior pays. 🇺🇸 America: Separate checks, please. At first glance, this looks like a simple restaurant scenario. But underneath it reveals something much deeper: culture shapes behavior. In many Asian societies, paying the bill is not just a financial transaction, it is a reflection of generosity, respect, relationships, and social status. In China, insisting on paying can be a way of showing hospitality and strengthening bonds. In Japan, fairness and harmony often lead to carefully splitting costs so nobody feels burdened. In Korea, hierarchy and respect for seniors remain deeply embedded in social interactions. In America, individual independence is highly valued, making separate checks perfectly normal and socially accepted. None of these approaches are right or wrong. They simply reflect different values: • Relationship vs. individuality • Hierarchy vs. equality • Collective responsibility vs. personal autonomy • Social obligation vs. personal choice The lesson goes far beyond restaurants. When working across cultures, what seems “normal” to you may feel strange or even rude, to someone else. The most successful global leaders aren’t the ones who impose their own norms. They’re the ones who understand the invisible cultural operating systems behind people’s actions. Because in a connected world, cultural intelligence is becoming just as important as technical intelligence. Question: When dining with friends, what’s the norm in your country, and which approach do you personally prefer? 👇

“Effortless is a myth.” The most elegant champion in sports history just reminded us that greatness is built on grit, patience, and loving the grind. Even legends lose almost half their points l, what matters is how fast you let it go and swing at the next one. Play free. Life > the court. ❤️

Vala Afshar @ValaAfshar

17 hours ago

Life lessons from @rogerfederer (must watch) 1 Effortless is a myth 2 Belief in yourself has to be earned 3 Grit > Gift 4 Discipline is talent 5 Trust and loving the process is talent 6 You can do your best and still lose 7 Life is bigger than the court

3 41 155 15K 85

China leads humanoid volume today, just like EVs. But this race doesn’t have to be zero-sum. China is executing the EV playbook on robots: massive scale, aggressive production, and now ~90% of global humanoid shipments. Companies like Unitree and Agibot are shipping thousands while Optimus ramps up. Their manufacturing speed and cost discipline are impressive, no denying it. Tesla and the US bring strengths in software intelligence, real-world learning, and rapid iteration (the same edge that keeps pushing FSD forward despite fierce EV competition). Elon has openly called China Tesla’s biggest humanoid rival, respect where it’s due. Here’s the bigger truth: Robots don’t need a single winner. This isn’t zero-sum. A world with millions of capable humanoids, from any nation, creates abundance: higher productivity, lower costs for goods and services, and humans freed for higher-value work. Both sides can thrive. Even better? Partnership is possible. Global supply chains, shared standards, and collaboration on safety/AI ethics could accelerate progress for everyone. Competition drives innovation; cooperation scales the benefits. China’s hustle + Tesla’s vision (and others worldwide) = faster path to a robot-rich future that lifts living standards globally. Let’s compete hard on excellence, then collaborate on impact. The real victory is a more abundant world.

Elonogy @ElonogyX

23 hours ago

Elon Musk: “China is a manufacturing powerhouse next level. If you take refining of ore, I’d say roughly China does twice as much ore refining on average as the rest of the world combined Like, we actually do rare earth ore mining in the U.S., send the rock… we put it on a

13 69 265 17K 28

“Over $100k in student debt, only $2,000 in the bank, and one computer.”
— Elon Musk (1995) Yet he still dropped out of Stanford, slept on the office couch, coded the server himself, cold-called businesses, and ate hot dogs while building Zip2 from nothing. You don’t need perfect conditions to start. 
You don’t need rich parents, big funding, or zero risk. You just need to dare to dream, then back it up with years of relentless grit, stamina, and hard work. The biggest successes often begin in the smallest, messiest rooms. Stop waiting for the “right time.”
Start building anyway. The future belongs to those who begin with nothing but belief and unbreakable determination. 💪

Sierra @Sierra_rak

a day ago

1995: $100k+ in student debt. $2,000 in the bank. One computer. That’s how Elon Musk started.

1 3 26 3K 1

Never Give Up: Rejected 11 Times by NASA But His Wife Refused to Let Him Quit He stood there in the suit, looked them in the eye, and calmly corrected them: “This is my 12th attempt.” This powerful scene (from motivational edits of A Million Miles Away) captures one of the most inspiring real stories in modern achievement: José M. Hernández, a Mexican-American migrant farmworker who went from picking grapes in California’s Central Valley, moving seasonally with his family, changing schools constantly, to becoming a NASA astronaut who orbited Earth aboard Space Shuttle Discovery on mission STS-128.7 Hernández applied to NASA’s astronaut program 11 times and was rejected every single time over more than a decade. He didn’t complain or quit. Instead, he systematically closed the gap: earning a master’s in electrical engineering, logging over 800 pilot hours, running marathons, earning his scuba certification, learning Russian, and even volunteering in Siberia for the U.S. Department of Energy, all while being a dedicated father of five. Every major decision in his life was deliberately aligned with his goal. On the 12th attempt, NASA finally said yes in 2004. His wife, Adela, played a pivotal role, refusing to let him give up when the rejections piled up. In today’s competitive landscape, rejection is inevitable. The winners are those who treat it as data, not defeat. Who keep showing up better prepared. Who have a support system that demands persistence. Your “11 rejections” might be lost pitches, failed hires, content that didn’t land, funding rejections, or career setbacks. The game changes on attempt 12 or 15, or 20, when most others have already walked away. José’s journey proves: Breakthroughs live on the other side of persistence. Align your daily decisions with your biggest goals. Build real capabilities. And never underestimate the power of someone who won’t let you quit.

How to Use Claude to Build a Startup Most people use Claude to answer questions. A small number use Claude to build businesses. The difference isn’t the AI. It’s the system. Instead of asking random prompts every day, create a structured workflow where Claude becomes your startup operating system. Here’s the framework: 🔹 Validate the Idea Use Claude to analyze market demand, identify customer pain points, study competitors, and pressure-test assumptions before investing time or money. 🔹 Create Your Startup Plan Define your mission, business model, value proposition, pricing strategy, go-to-market plan, KPIs, and roadmap. 🔹 Build Your MVP Faster Turn ideas into product requirements, user stories, technical specifications, wireframes, and code. 🔹 Build and Launch Create landing pages, marketing assets, sales scripts, email sequences, customer onboarding flows, and operational processes. 🔹 Get Users and Traction Develop your ideal customer profile, messaging, content strategy, outreach campaigns, and growth experiments. 🔹 Analyze and Improve Review customer feedback, identify bottlenecks, track metrics, and continuously improve your product and business. 🔹 Fund and Scale Generate investor decks, financial models, due diligence documents, operating manuals, and growth playbooks. 🔹 Automate and Operate Build AI-powered workflows, automate repetitive work, document SOPs, and create systems that scale without increasing headcount. The biggest opportunity of this decade isn’t AI itself. It’s founders learning how to combine AI, systems, and execution. The startup of the future may not have 50 employees. It may have 5 people and an army of AI agents. That’s why I no longer think of Claude as a chatbot. I think of it as a Startup Operating System. What part of your business would you want an AI co-founder to help you run? 👇 Let me know below. P.S. Sign up for the free AI Diagnostic and newsletter at 10xme.biz.

JAPAN JUST FIRED A SHOT IN THE AI WAR. For the last 2 years, the AI race has looked like a two-horse race: 🇺🇸 United States 🇨🇳 China Meanwhile, most people assumed Japan was watching quietly from the sidelines. Not anymore. Sakana AI (@SakanaAILabs) just released Fugu Ultra, and according to its benchmark results, it’s going head-to-head with some of the most advanced AI systems from OpenAI, Anthropic, Google, and other frontier labs. On several coding, reasoning, and scientific benchmarks, Fugu Ultra is matching or even surpassing, the scores of leading models. But here’s the part that caught my attention: Sakana didn’t win by building a bigger model. They won by challenging the assumption that one giant model is the future. Instead, Fugu uses a multi-agent orchestration system, a conductor that coordinates multiple frontier models and specialized agents to solve problems collectively. Think less “super genius” and more “elite task force.” This is a fascinating shift. The next breakthrough in AI may not come from building a larger LLM. It may come from building a better AI operating system that knows: • Which model to use • When to use it • How to coordinate multiple agents • How to combine their outputs into a superior result In other words: The future may belong to orchestration, not just intelligence. This is especially interesting because it validates what many of us building AI systems have suspected: The real moat isn’t the model. The moat is the architecture. Japan’s entry into the race also sends a broader message: Never assume innovation only comes from Silicon Valley or Beijing. The next AI breakthrough could emerge from Tokyo, Singapore, Seoul, Paris or a startup you’ve never heard of. The AI race is no longer USA vs China. It’s becoming everyone vs everyone. And that’s great news for innovation. What’s your take? Is the future a single supermodel… Or a swarm of specialized AI agents working together? Source : sakana.ai/fugu-release/

Most people are using Claude like it’s ChatGPT. That’s like buying a Formula 1 car and driving it in a parking lot. After testing virtually every major AI model and workflow over the last year, I’ve noticed something: 👉 The difference between average AI users and power users isn’t the model. 👉 It’s the SYSTEM. Most people: ❌ Open Claude ❌ Type a prompt ❌ Get an answer ❌ Start over again tomorrow Power users: ✅ Build reusable workflows ✅ Create project knowledge bases ✅ Store context once and reuse forever ✅ Turn AI into a digital operating system A few game-changing lessons from this cheat sheet: 1️⃣ Claude isn’t one tool. It’s 5. Chat, Projects, Artifacts, Skills, and Connectors each serve different purposes. 2️⃣ Context is the new prompt. A well-structured project folder beats 100 clever prompts. 3️⃣ Build once. Reuse forever. Create templates, standard operating procedures, and AI playbooks that compound over time. 4️⃣ Stop prompting. Start operating. The biggest productivity gains come when AI understands your business, goals, voice, and workflows. 5️⃣ Your AI should become your second brain. Not another search engine. The future belongs to people who know how to orchestrate AI, not just talk to it. I’ve expanded this Power Claude User Cheat Sheet to help executives, founders, consultants, and professionals get significantly more value from AI. 👇 What’s your biggest challenge with Claude or AI right now? P.S. If you’d like a free AI diagnostic and AI strategy newsletter, visit 10xme.biz.

Probably the best ad you'll watch this Father's Day. I wasn’t expecting to cry today, but this ad got me. 😍

A special final flight for a special dad!❤️ Credit : Southwest Airlines

In a noisy, hurried world, Sonna stopped everything to offer more than free coffee, she offered wisdom wrapped in kindness. “Be good to the next person… because you never know when you’ll meet again.” This is what we can learn from this kind-hearted soul: • Kindness is never wasted. • Empathy is the real currency of life. • One genuine interaction can heal, inspire, and ripple outward forever. • You don’t need a platform to change hearts, just an open heart. Thank you, Sonna. You reminded us that the smallest acts, done with love, are revolutionary.❤️

Massimo @Rainmaker1973

2 days ago

A beautiful example of humanity.

62 630 3K 56K 137

The AGI Race Isn’t About Technology, It’s About Geopolitical Power. And the Clock Is Ticking. I just watched Ashton Kutcher lay it out with striking clarity: “We are in a war… between the free world and China… We’re about a thousand days away from having one of these models go hyperbolic. It starts self-training… That model will be able to do about 25,000 years of thinking in one second. That’s the most powerful tool humanity has ever known.” This isn’t sci-fi speculation. This is a sober recognition that the first nation or bloc to achieve AGI will secure decisive advantages in economics, defense, intelligence, innovation, and global influence. The incentive structure is brutal: move faster, even if society isn’t fully ready. The real risk isn’t AGI itself. It’s that raw competition forces premature, uncontrolled deployment. Yet hesitation is not a viable strategy. The race is on. China is all-in. The question for leaders, founders, and strategists is whether we will be spectators or active shapers of what comes next. Why this matters for executives and builders right now: • A model capable of 25,000 years of thinking in one second doesn’t just automate tasks. It compresses decades of R&D, strategy, and problem-solving into hours. • This creates asymmetric opportunities for lean, AI-native organizations to outmaneuver legacy players. • The winners will be those who have already built robust agentic workflows, proprietary data moats, and execution muscle before the hyperbolic phase hits. In my work, I’m seeing a clear pattern: the leaders pulling ahead are treating AI as a strategic operating system today, not a future experiment. Asia’s position is uniquely powerful. We combine talent density, pragmatic regulation, and leapfrog ambition. But only those who act with urgency will capture the upside. The next 1,000 days will define the next century. Leaders: What is your boldest move to prepare? Are you building agentic systems now? Stress-testing your organization for intelligence explosion? Drop your thoughts in the comments. Let’s exchange practical insights, no hype, just executable strategy. The most powerful tool in human history is arriving. The only question is whether you’ll wield it. Credit: Ashton Kutcher

Most people use Claude like a search engine. Stanford is teaching AI to research like a PhD. The framework is called STORM (Synthesis of Topic Outlines through Retrieval and Multi-perspective Question Asking). Instead of asking AI for answers… STORM teaches AI to investigate. The process: ① Generate Perspectives “Who are the experts, critics, practitioners, stakeholders, and opposing voices?” ② Ask Better Questions “What questions would each perspective ask that others would miss?” ③ Retrieve Evidence Gather primary sources, contradictory evidence, research papers, data, and expert opinions. ④ Synthesize Findings Identify patterns, disagreements, uncertainties, and emerging trends. ⑤ Challenge Conclusions “Do not optimize for agreement. Optimize for truth. Find evidence that disproves your conclusions.” This is where things get interesting. Most AI users are trying to generate content faster. Researchers are using AI to compress months of research into hours. Instead of getting: ❌ A polished opinion You get: ✅ Multiple viewpoints ✅ Evidence-backed analysis ✅ Contradictions and blind spots ✅ Knowledge gaps ✅ Future research directions ✅ Executive-level insights Want to see it in action? Stanford has made a live version publicly available. You can try it at: storm.genie.stanford.edu No signup required. Just enter a topic and watch it build a fully sourced research article in front of you. Even more impressive: Stanford open-sourced the entire project. The full code is available at: buff.ly/bB9yRn2 MIT licensed. You can run it locally on your own laptop, modify it, or integrate the workflow into your own AI stack. We’re entering a world where everyone has access to research capabilities that previously required teams of analysts, consultants, and academics. The competitive advantage is no longer access to information. It’s knowing how to investigate better than everyone else. Next time you open Claude, ChatGPT, Gemini, or DeepSeek… Don’t ask for an answer. Ask it to investigate. P.S. If you’re wondering where AI can create the biggest impact in your business, get a free AI diagnostic and join the 10xme newsletter at 10xme.biz.

One bad apple tried to ruin the moment. Mexico responded with class, accountability, and pure heart. They welcomed her, apologized personally, gifted a signed national jersey, and showed the world what hospitality really means. Her smile in that video says everything. This is the true magic of the World Cup: one person can cause harm, but an entire nation can heal the wound. Football > hate. Always. 🇰🇷❤️🇲🇽

Kara @UTDKarra

2 days ago

The South Korean lady who was racially abused by a Mexican fan was welcomed by the Mexican govt and received an apology for what happened 🇰🇷❤️🇲🇽 The Mexico FA also gave her a signed jersey & took pictures with her. This is what the World Cup is all about

525 5K 74K 1.5M 3K

This is NOT a broad bull market, it’s an AI supercycle wearing the S&P 500 as a mask. AI stocks now ~47% of the entire index (up from 27% in early 2023). Just 84 AI names + 22 energy stocks did all the heavy lifting while everything else bled -$1T. The index hits new all-time highs on razor-thin breadth. Momentum is strong and AI’s long-term story is powerful but this level of concentration is extremely risky. A sharp 10-15%+ pullback isn’t guaranteed tomorrow, but the setup is fragile. Any stumble in AI earnings, capex slowdown, or sentiment shift could hit hard and fast. High-conviction AI bet disguised as a market rally. Size positions carefully and watch those key earnings closely. What do you think, melt-up continues or rotation/correction coming? 📉🚀

The Kobeissi Letter @KobeissiLetter

3 days ago

The market is being driven by just 2 sectors: The S&P 500 has added over +$5 trillion in market cap so far in 2026. Meanwhile, AI stocks have added +$6 trillion in value, followed by +$200 billion added by the energy sector. At the same time, other sectors have erased -$1

149 402 2K 353K 441

This chart perfectly captures what many people in the West are still underestimating about China’s AI strategy. The narrative has largely focused on who has the most powerful model. But the more important question may be: Who can deliver intelligence most efficiently? China’s leading AI companies appear to be pursuing a fundamentally different approach from many frontier labs: → Less experimentation, more optimization → Less brute-force compute, more engineering efficiency → Less proprietary isolation, more ecosystem learning What stands out is that they’re not just optimizing model training costs. They’re attacking the entire cost stack: • Model architecture • GPU utilization • Inference efficiency • Infrastructure engineering • Energy consumption • Open-source knowledge transfer This mirrors what we’ve seen repeatedly across other industries. China doesn’t always win by inventing first. It often wins by making something dramatically more efficient, scalable, and affordable. The same pattern may now be emerging in AI. The first phase of the AI race was about capability. The next phase may be about economics. And economics has a habit of reshaping entire industries. The question isn’t whether a model is 5% smarter. The question is whether it can be delivered at 50% of the cost. That’s where things get interesting.

zerohedge @zerohedge

3 days ago

Chinese model efficiency snapshot: inference and training

15 57 292 103K 146

This chart is one of the cleanest examples of a true paradigm shift I’ve ever seen. 60 years of flatline, Space Race, Apollo, Shuttle, ISS, endless government billions, nothing moved the needle past ~120 launches/year. Then reusability hit. One company cracked the code on landing and refllying boosters, and the entire game changed overnight. 2025: record ~321 global launches. The vertical spike is almost entirely SpaceX. Strip them out and the world is still stuck on the old plateau. Legacy players didn’t fail from lack of effort. They failed because they kept building disposable rockets in a world that suddenly didn’t need them. Private innovation didn’t just add launches, it rewrote the economics of orbit and opened the door to what’s next. The plateau is dead. The reusability era is just getting started. 🚀

Rand Group @randgroup

3 days ago

From 1957 to 2020, humanity launched roughly 60 to 120 rockets a year into orbit. For sixty years, the number barely moved. The Space Race didn't break it. The Shuttle program didn't break it. The ISS didn't break it. Billions in government funding, Cold War urgency, national

5 15 138 17K 5

Eric Schmidt’s: “China’s AI is largely open source… and therefore largely uncontrolled by us.” Schmidt sees open source as a problem because it threatens control. I see it as the solution. Just like Linux didn’t kill proprietary OS, it complemented Windows and macOS, powering the internet, clouds, supercomputers, and Android while driving down costs, boosting security through transparency, and making computing accessible to the entire world. Open-source AI can do the same: •No vendor lock-in or per-token taxes •Full transparency and auditability •Customizable for every language, culture, and use case •Sovereign, nations, companies, and individuals can actually own their intelligence stack •Faster collective innovation through “many eyes” Proprietary models will still thrive where polish and enterprise support matter. But open models provide the open foundation that prevents any single player (East or West) from owning the future. Control vs. capability.
Hegemony vs. human flourishing. The open path is clearly the better one.

Arnaud Bertrand @RnaudBertrand

2 days ago

Eric Schmidt saying the quiet part out loud: "What I don't like about [China's AI] is that it's all open source which means it's largely uncontrolled and not controlled in any way by us." He adds, "if that makes you feel any better," that only 2 or 3 countries can be independent

512 3K 10K 843K 2K

Happy Father’s Day to all the fathers and father figures in my network and among my followers.  This video montage captures the raw emotion of some of the world’s most accomplished people, Oscar winners, actors, directors, standing on the biggest stages, pausing amid glory to credit their dads. It’s a powerful reminder that behind every success story is often a father who showed up, taught resilience, modeled strength, instilled belief, and opened doors to the world. To every dad reading this, your impact compounds far beyond what metrics can measure. The late nights, the tough lessons, the belief when no one else saw it, the sacrifices that let your kids chase bigger horizons. That’s the real legacy. To my followers who are fathers: Today, take a moment to feel seen. Your work, your leadership, your presence matters more than the algorithms, valuations, or headlines. Keep leading by example. The next generation is watching and will one day stand on your shoulders. If you’re a dad, reply with one lesson your father taught you (or one you’re passing on). Tag a father who inspired you. Let’s flood the feed with respect. Video credit: TikTok @ surgemotions