Yax Sheth

Throwback to Crypto Expo Dubai! I have met 100+ interesting Web3 people on that day, 20+ potential Web3 Projects! I am consulting 3 Web3 start ups met in Crypto expo Dubai! Super bullish for Web3 💪🏻

Birthday Morning be like! ❤️🔥🫶🏻

Iran just turned the Strait of Hormuz into a Bitcoin toll booth. $1 per barrel of oil. A fully loaded supertanker? That's $2 million. Payable in Bitcoin. Seconds to pay. No SWIFT. No tracing. No freezing. A sanctioned nation controlling 20% of the world's oil supply is now collecting crypto tolls at the most important chokepoint on earth. Their parliament already passed it into law. Everyone's sharing the viral math. 130 ships a day, 3,600 BTC daily, 8x Bitcoin's entire mining output. The math is right but the context is wrong. The strait is still 90% closed. 800+ ships are stranded. Maersk won't touch it. But that's not the point. The point is a country just proved you can run sovereign revenue collection on Bitcoin infrastructure. Outside the dollar system. At gunpoint if needed. Even at 10% of normal traffic, Iran pulls in $20M+ a day in crypto. Trump floated a "joint venture" with Iran on the tolls. Read that again. The petrodollar didn't die today. But it just got a preview of its funeral. Now imagine what happens when other nations start taking notes. #Bitcoin #Crypto #Geopolitics #StraitOfHormuz #Web3

India’s got Latent Season 2 🔥

🚨 Oracle just fired 30,000 people via a 6 AM email. The stock went up 5%. Let that satisfying little irony wash over you. Oracle, a company that just reported $6.1 billion in GAAP net income last quarter (up 95% year over year), woke up 30,000 employees before sunrise on March 31st with an email signed “Oracle Leadership.” No phone call. No meeting. No heads up from their manager. Just a cold paragraph telling them their role has been “eliminated as part of a broader organizational change” and today is their last working day. Some of these people gave Oracle 10+ years. Some had stock vesting in June, just 2 days away. Forfeited. Immediately. The reason? Oracle needs $8 to $10 billion in free cash flow to fund $156 billion in AI data center buildout. They’ve taken on $58 billion in new debt in two months. Their stock has lost more than half its value since September 2025. So the solution wasn’t to slow down. Wasn’t to recalibrate. It was to delete 18% of the workforce and redirect their salaries into GPU racks. And Larry Ellison, Oracle’s chairman, CTO, and a man worth over $200 billion, owns 98% of a Hawaiian island he bought for $300 million. He owns the grocery store. The gas station. The cemetery. The hotel. The newspaper. He literally owns the town. That man decided 30,000 families can figure it out. But this isn’t an Oracle problem. This is the playbook now. → IBM cut 7,800 and replaced them with AI in 2023. → Amazon laid off 27,000 while reporting record revenue. → Meta has been doing rolling layoffs while printing money. → Google dropped 12,000 while sitting on $100 billion+ in cash. → Block just cut roughly 40% of its workforce and the stock surged. Every single time: record profits, mass layoffs, stock goes up. They told you to learn to code. You learned to code. They told you to upskill. You upskilled. They told you to “add value.” You added so much value that net income went up 95%. And then they replaced you with the thing you helped build and sent the termination letter before your alarm went off. The company made record profits. And the reward for the people who made it happen? An email at 6 AM and a forfeited stock vest. Meanwhile, Wall Street celebrated. Oracle stock jumped 5% on the news. Because in this economy, your termination is someone else’s alpha. This is not innovation. This is extraction. And if you’re sitting at your desk today thinking “at least it’s not me,” understand this: you’re not safe. You’re just not expensive enough yet. Stay sharp. Build your own leverage. Because loyalty to a corporation is a one way street and they just proved it. Again. What’s your take? Is this the new normal, or are we going to start demanding better? #OracleLayoffs #AI #TechLayoffs #FutureOfWork #CorporateAmerica #Leadership

Most people caught 5 of these releases. There were 73. Anthropic shipped 73 product releases in 52 days. Not announcements. Not blog posts. Actual shipped product: Claude Code, Cowork, the API, models, all of it running in parallel. Look at the calendar in this image. February had some breathing room. By March 9, it became almost every single day. Back to back. Here’s the part that breaks my brain though. Anthropic’s own engineers use Claude for 60% of their work now, up from 28% a year ago, reporting a 50% productivity boost. They’re not building these products in spite of AI. They’re building them BECAUSE of AI. Cowork was built using Claude Code in 10 days. The tool that ships the tools is the tool itself. That’s a loop most companies haven’t even begun to think about. Claude Code’s annualized revenue topped $1B by end of 2025 and had already doubled to $2.5B by February. That’s not a research project anymore. That’s a platform. Every AI discussion I have with founders, investors, builders across Web3 and beyond, it always circles back to who’s actually shipping. Not who’s raising. Not who’s announcing. Who’s SHIPPING. The answer right now is pretty clear. It’s not about picking the best AI tool. It’s about recognizing when one platform’s shipping velocity makes the decision for you. What are you actually building with these tools right now? Curious what the builders here are seeing on the ground. #Anthropic #AIAgents #Web3 #Builders #ArtificialIntelligence

OpenAI didn’t kill Sora because it failed. They killed it because something bigger is coming and they can’t afford distraction. Sam Altman told employees internally that ending Sora frees up resources for next-generation AI models. That’s not the language of a company cutting losses. That’s the language of a company making a calculated bet. Think about what Sora actually was. A video app that went viral, hit 1 million downloads in under 5 days , and then… slowly faded. Because viral ≠ valuable. The compute cost was enormous, the copyright drama was endless, and the monetization was essentially zero. Meanwhile OpenAI needs hundreds of billions. For GPUs, infrastructure, talent, data centers. Every dollar of compute going into Sora is a dollar NOT going into ChatGPT, Codex, or whatever they’re building next. They’re combining ChatGPT, Codex, and their browser into a single desktop super app. That’s the real move. Not a flashy video tool. A platform people use every single day for work. The Disney partnership is done too. A $1B deal, gone. When you’re willing to walk away from a billion dollar partnership to stay focused, that tells you everything about how serious the next phase is. Most people will write this story as “OpenAI failed at video.” The real story is: OpenAI is preparing for something that makes Sora look like a side project. What do you think they’re actually building toward? I’m curious what people outside the hype cycle are reading into this. #OpenAI #AIStrategy #Web3 #Crypto #ArtificialIntelligence

Claude on mission! They are publishing new updates every single day!

@steipete Claude Code Plugin will be game changer!

Thinking how we can evolve openclaw plugins to be more powerful while also making core leaner. Also wanna add support for claude code/codex plugin bundles. Good stuff coming soon!

@steipete @chrysb 🤣🤣🤣

@chrysb what whacky model are you torturing your claw with?

my claw got drunk last night and spent $0.13 trying to run a script and said HELLO_OK lol. we need deterministic cron!

@steipete @morganlinton @denisyarats Hahaha! You made every billion dollar company to change their view point

@morganlinton @denisyarats Wonder where I seen this idea implemented before.

The cofounder and CTO of Perplexity, @denisyarats just said internally at Perplexity they’re moving away from MCPs and instead using APIs and CLIs 👀

@0xsachi Yes! Any day…

is claude max worth it?

@amarifields_ Interested

would anyone be interested in a group chat for founders trying to fundraise to share tips, resources, and what is actually working comment below if you would join

@raagulanpathy Congratulations 🤝🥳

“To the dreamers” Today we announce KAST has raised a US$80M Series A, co-led by QED and Left Lane. We launched in July 2024, and initially there was no belief in my thesis that the biggest opportunity wasn’t stablecoins themselves, but building a global-first financial platform on top of stablecoins. But one step at a time, this changed. Many individuals came into the pre-seed, and Peak XV and Hong Shan led the seed round. Now QED and Left Lane double down with conviction. Yes it’s a historic day; for a company to go from zero to such a large raise in ~16 months is rare. But for the people at KAST, the ones who I really need to thank for showing up 24x7 - it’s back to the grind. We know the only thing that matters, is not flashy raises, but delivering an outstanding product and service to the billions of people who need it, and the customers who trust us everyday. We are not perfect, we are not even two years old, but god damn we try our absolute best. My final points are to the dreamers and believers. Building KAST, the mission was simple, be the stable backbone to people’s financial life, so they can fly as high as they want. KAST is built for the dreamers, and KAST is for believers who trusted us from very early. People like me, a simple dreamer. KAST is so personal to me, because it came from 25 years of trying and failing in life. I built so many side projects, which never went anywhere. I tried so many startups, some which even scaled and then failed. My parents are now in their late 70s, I thought they’d never see my succeed in the one thing I wanted to do the most, which was to build a generational company which made everyone proud - those who invest in KAST, work at KAST, and those who use KAST. Whilst we are not there yet, in fact far far from it, I’m evidence that if you just stay crazy and keep believing, and keep rolling the dice… eventually, you might get the momentum for it to happen. I believe KAST is in the fullness of time, a trillion dollar company for a billion+ people. I’ve never sold a share, and quite the opposite in 2025 kept buying in a lot more. My conviction is infinite on this one. People will think I’m stupid to go so deep. I don’t care. This is beyond money. Thanks for all the incredible support. Now back to work. Peace ✌🏾

India are World Champions. Again. 🇮🇳🏆 Three time T20 World Cup winners. First team ever to defend the title. This one is historic. But what stuck with me wasn’t the trophy. It was HOW they got there. Abhishek Sharma walked into this tournament as the ICC’s number 1 ranked T20 batter. But his form through the early matches wasn’t there. The runs weren’t coming. Most teams would have panicked. Most leaders would have shuffled. Suryakumar didn’t. The management didn’t. They backed him. And in the final, Abhishek scored the fastest fifty of the entire tournament. 18 balls. That’s what backing your people looks like. But it wasn’t just Abhishek. Every single match, someone different stepped up. Suryakumar rescuing India against USA almost single-handedly. Ishan Kishan and Hardik going berserk against Namibia. Chakaravarthy picking wickets when India needed breakthroughs. Different player, different game, same result. That’s not talent. That’s culture. Building a company works exactly the same way. Building a brand works exactly the same way. You don’t win because one person is extraordinary. You win because your environment makes anyone capable of stepping up on any given day. Back your people before they prove it. That’s what great leaders do. Jai Hind 🇮🇳 #TeamIndia #T20WorldCup #Leadership #Entrepreneurship #Web3

@jbarbier Using Claude Cowork is super addictive for me

Using Claude Code has a weird side effect: You don't just get more productive, you actually want to work more. There's something addictive about watching a product being born in real time in front of your eyes. "One last feature" after "one last feature" and it's already past 3am.

Full conversation with Harsh Agarwal on YaxVaani 👇 youtu.be/2DxOgBpvaCw

Meet Harsh Agarwal @HarsAgarwal > At 17, he dropped out of school and started mining Bitcoin from his bedroom. >By the time college started, he was already making Rs.1 lakh a month. > He didn't stop there. > Cracked CA and CFA while trading crypto on the side. > Got into JP Morgan. Then moved to DeFi. > Became part of Kyber Network when it got hacked for $100M. > Helped make the ethical call that saved the team's integrity. > Then Polygon came calling. > Meetings with Jio. Tata. Tarak Mehta. > Today he is Investment Director at Cypher Capital, Dubai. > Backed Solana, SUI, SEI, and Hyperliquid before they were obvious calls. > Connected to Phoenix Group, the 6th largest Bitcoin miner on earth. > He spends $1000 every month on AI tools. > Because time is the only thing money cannot return. > And after all of this, the crores, the exits, the deals, he told me: "Main paise ke liye nahi bhaag raha. Social capital banana hai."

@VadimStrizheus Claude CoWork

As a founder, which AI tool are you using in 2026? 1. OpenClaw 2. Perplexity Computer 3. N8n 4. Claude Cowork 5. Manus

@jessepollak Maaan! I thought i am the only one having addiction with building agents

help I'm addicted to building with agents

@nicbstme Claude code & Cowork winning over openclaw

Why do I need openclaw? Am I missing out? I run Claude Code from my phone with /remote-control. It has access to my filesystem, professional and personal Drive and emails with GogCLI, my sms etc. It seems like 80% of the value prop without the security headache and token cost.