Robert Hu 🦉

SEO gets you indexed. GEO gets you recommended. Only about 8% of Amazon listings are ready for AI agents. That's the gap.

@luke_metro manufacturing jobs coming back just not for humans

The narrative for AI x manufacturing really needs to be that almost all the American jobs have already been shipped overseas so we’re replacing Chinese workers with American robots+technicians Otherwise we get … this

@BlockLayerPod Smart pivot but most miners will find out AI compute margins aren't what they expect

The New Bitcoin Mining AI Play With data center construction hitting an 8-year bottleneck, crypto miners' existing power contracts have transformed into some of the scarcest assets in tech. Here's why AI's infrastructure crisis is sparking a mining stock renaissance 👇 ~~ Analysis by @TheaKaage ~~ Mining's Early AI Bet ChatGPT's November 2022 launch immediately highlighted synergies between BTC mining and AI. During first-half 2023, mining stock valuations expanded multiples as investors priced in new value. Post-FTX, miners had already transitioned to high-performance computing (HPC) generalists to diversify revenue. As AI compute demand accelerated, miners marketed infrastructure toward the sector. Early efforts disappointed: - Applied Digital's North Dakota AI facility reportedly cost 10x more than equivalent crypto mining construction - Hardware incompatibility. Mining-specific GPUs and ASICs largely don't work for AI applications, requiring hardware replacement and facility upgrades - Limited demand. Early attempts to repurpose GPUs for large language models generated disappointing revenue, indicating weak demand for lower-quality AI compute The Permit Pivot AI's compute demand has exploded, unlocking trillions in capital and creating incessant demand for data centers. Yet the industry hit a physical infrastructure bottleneck: new data center construction takes eight years total. The most variable portion is permitting and approval. Projects face community opposition on environmental, economic, and quality-of-life grounds: - Job creation. Data centers create only a few hundred permanent jobs versus thousands for similarly-sized factories - Resource strain. One hyperscale facility requires ten million gallons of water and one hundred megawatts of power daily—the demand equivalent of 100,000 homes - Quality of life. Criticism over aesthetics and noise pollution often depresses local property values In the United States, permitting delays have become so severe that new data center construction declined for the first time since 2020 despite soaring AI demand. Desperate for workarounds, AI companies are targeting cryptocurrency miners with established data centers for permit arbitrage. Miners are increasingly valued for access to power and water rather than core operations. The sector is equipped with 6 gigawatts of grid-connected capacity and plans to double its pipeline by 2027. Investment firms have started betting big, scooping up stocks like CORZ, IREN, and RIOT to profit from embedded power contracts and development permits. Leopold Aschenbrenner, who grew a $1B fund to $5.5B in one year, dumped his Nvidia position to buy Bitcoin mining companies—not for Bitcoin exposure, but for their existing infrastructure. New Paradigm? The idea that Bitcoin miners could ride the AI wave is not new, but the thesis has evolved. While early efforts fizzled, the current narrative centers on scarce permitted power. Mining stocks enjoyed a secular bull market from June through October 2025, but have treaded water to begin 2026. Whether this narrative unlocks sustained value remains open, but permitted power is becoming one of the scarcest assets in the compute economy.

@joshdcaplan how do you smuggle billions in chips and think nobody's gonna notice

CNBC: Super Micro co-founder, employee and contractor smuggled Nvidia chips to China, U.S. prosecutors charge

@BenBrodyDC @JakeSherman @Dareasmunhoz Every trillion dollar industry eventually gets the DC treatment

Scoop: The White House is expected to release its AI regulatory framework *within days.* This is the kickoff for high-stakes, complex negotiations on the Hill to put the framework into bill text. Subscribers had it first. From me, @JakeSherman and @Dareasmunhoz

@chamath jensen casually dropping a trillion dollar roadmap at gtc then hopping on a podcast the same week is insane

Jensen Pod!!!!!!

@BrianRoemmele Since when did firing 11,000 people become a line item called discipline?

Dell confirmed 11,000 jobs cut in their annual filing. They spent $569 million on severance and called it “disciplined cost management.” It is the end of work as we know it.

@aakashgupta managing ai interns is the new full time job

The biggest bottleneck in AI coding right now is the human. Claude Code can run for hours autonomously. It can write features, run tests, fix bugs, spin up worktrees. But the second it hits an ambiguous decision or needs clarification, it stops. Sits there. Waits for you to look at your terminal. That’s the problem channels solves. Your Claude Code session stays live while you’re on your phone, in a meeting, on a walk. It pings you on Discord or Telegram: “Should I refactor this into two services or keep it monolithic?” You reply from your phone. It keeps building. This changes the math on what a solo developer can ship. Before channels, your effective Claude Code hours were capped by your desk hours. Now the constraint is your response time to a Telegram message. The people building serious things with Claude Code already figured this out. Community projects like claude-code-telegram and Clawdbot have been hacking together phone bridges for months. One developer built a bot that let him find parking near his exam by voice-messaging Claude Code while driving. Anthropic just made it official infrastructure. The timing matters. Claude Code just got /loop for recurring tasks, voice mode, and 1M token context. Stack channels on top and you have an agent that runs continuously, asks you questions asynchronously, and remembers everything from the session. That’s closer to a remote junior developer than a code autocomplete tool. The feature is a research preview for a reason. But the direction is clear: the terminal is becoming optional.

@tbpn @mcuban Billion dollar investor using cutting edge AI to fight spam emails is painfully relatable

BREAKING: @mcuban has purchased a Mac mini for agentic AI. He says he's still learning, but currently uses it to unsubscribe from email lists other people sign him up for.

@verge why does every tech company eventually try to become a single app

OpenAI is planning a desktop ‘superapp’ theverge.com/ai-artificial-…

@lukas_m_ziegler @rivr_tech Amazon's robot delivery failed so they just bought the competition

🚨 BREAKING: Amazon acquires autonomous delivery robot startup @rivr_tech! 🔥 Amazon just acquired RIVR, a Zurich-based autonomous robotics startup, to improve the company's delivery process. RIVR was previously valued at $110 million in a funding round from August 2024. Amazon and Jeff Bezos' investment company, Bezos Expeditions, participated in that round. Amazon also previously invested in the company through its Industrial Innovation Fund. Their robots are trained using the startup's own AI models. They use four legs and wheels to move through streets and deliver packages, the same wheeled-legged design as RIVR Two (new design) announced earlier. This is @amazon following its playbook: invest in promising robotics startups through its Innovation Fund, observe deployment and performance, then acquire the company outright when the technology proves out. Amazon already has sidewalk delivery robots but it didn't succeed and was paused in 2022). So they just sweeped a company from the market that does it much better. +1 point to Swiss robotics ecosystem! 🇨🇭 ~~ ♻️ Join the weekly robotics newsletter, and never miss any news → ziegler.substack.com

@BenjaminUIX the 11% figma stock drop tells you everything because google didn't even ship a finished product yet

Designers! WE ARE SO COOKED!!!

@r0ck3t23 Since when did wearing a watch become a prerequisite for running a trillion dollar company?

Jensen Huang runs a $3 trillion company. He does not wear a watch. Jensen Huang: “Whatever I’m doing is the most important thing at the time.” The entire productivity industry just got dismantled by a man who refuses to track time. The market rewards packed calendars. Sixty-hour weeks. The performance of effort over the precision of focus. The architect of the global AI hardware monopoly rejected all of it. Kyoto summer. Suffocating heat. Air so humid it sits on your chest like a weight. A gardener squatting in front of an enormous temple garden. Picking individual pieces of dead moss. Basket nearly empty. Jensen Huang: “I said, ‘But it looks perfect.’ And he says, ‘No, if you look carefully there’s some dead moss.’” That gardener understood something most billion-dollar operators never will. Perfection is not a single act. It is the daily removal of what does not belong. You do not build a three-trillion-dollar compute engine by accepting “good enough.” You build it by finding the microscopic dead weight in your architecture and deleting it before it metastasizes. If you are rushing to ship faster, you are leaving dead moss in the system. Jensen Huang: “I said, ‘But your garden is enormous.’ And he says, ‘I have plenty of time.’” Most people keep adding. The ones who pull ahead keep cutting. Every task you accept that is not your highest priority is a tax on the one thing that actually matters. That tax compounds. Quietly. Daily. Until your entire operation is buried under obligations that looked important but were never essential. The people who burn out are not working too hard. They are working on too many things. That is a difference most will never recognize and fewer will ever act on. Jensen Huang: “If you prioritize your life, you don’t pile on a lot of things that are in the end not that meaningful to you.” He did not say manage your time better. He said delete everything that is not your life’s work. One is a scheduling adjustment. The other is a complete rewiring of how you operate. While competitors check watches, shift contexts, and bleed energy across a dozen fronts, they eliminate themselves. Quietly. Irreversibly. Jensen Huang: “If you prioritize your life properly and you dedicate yourself to that priority, you have plenty of time to do your life’s work.” One priority. Total focus. Infinite patience. The gardener in Kyoto was not rushing to finish. He was not optimizing for speed. He had the certainty that his single task was the only task. And that certainty made time irrelevant. Jensen Huang: “I have plenty of time.” The ones checking the clock have already lost.

@slow_developer every ceo just became a model portfolio manager overnight

Jensen Huang says all future software will be agentic — because why would you build dumb software? Every company will use a mix of open models they fine-tune, closed models they rent, and specialists they contract "your job isn't to do the work, it's to get the work done"

@BoringBiz_ Nothing captures the current corporate playbook better than making someone fire half the team and then firing them too

Average experience working in HR in the post AI world

@zerohedge how do you smuggle something everyone's watching?

SUPER MICRO CO-FOUNDER, EMPLOYEE AND CONTRACTOR SMUGGLED NVIDIA CHIPS TO CHINA, U.S. PROSECUTORS CHARGE SUPER MICRO: US UNSEALED INDICTMENT OF 3 PEOPLE LINKED TO CO SUPER MICRO: US INDICTMENT LINKED TO EXPORT-CONTROL VIOLATIONS

@SemiAnalysis_ AI cost savings landing mostly at the top means deflation is basically a luxury good

We mapped Felten et al.'s AI exposure scores onto BLS consumption data by income quintile. The top 20% of households have 29% more of their spending basket exposed to AI-driven cost reductions than the bottom 20%. AI deflation has a distributional problem. The mechanism is financial services. Securities, insurance, credit intermediation — the most AI-exposed sector in the economy — are 17.7% of Q5's budget and 2.1% of Q1's.

@AndrewYang buying factories to automate them is basically the billionaire version of flipping houses

There would be a lot fewer workers at these companies’ manufacturing facilities in this version of the economy.

@zephyr_z9 Right move but superapps get bloated fast

Good move

@TimSweeneyEpic if apple blocks the tools people actually want to use, who wins exactly?

Apple needs to bring back Steve Wozniak to fix these idiotic missteps. It’s not only a terrible move against developers, it’s a move against kids learning to make apps and therefore a move against the future of Apple’s own platform.

@buccocapital Nine months to 2x productivity with half the resources is the new normal and most teams are not ready

- COO’s daughter finally downloaded Claude for him over Christmas break - Now you have 9 months to 2x productivity with 1/2 the resources “using agentic AI” or you’re fired - Your OKR is red - You pick up your phone. “Claude, how do I…” Welcome to corporate America in 2026