In Today’s Issue:
🤑 NVIDIA's Blackwell Ultra delivers 50x better efficiency, crushing token costs for the "Agentic AI" era.

💵 The $10B Indiana Hub: Meta breaks ground on a massive 1GW data center, pledging millions in local energy relief. 📈 GLM-5's Rise: China’s new open-source heavyweight matches Claude 4.5 and GPT-5.2 in real-world engineering.

📉 Labor’s Shrinking Share: New data shows a $2 trillion wealth shift from workers to capital as AI profits surge.

✨ And more AI goodness…

Dear Readers,

NVIDIA's Blackwell Ultra just landed benchmark results that make Jensen Huang look like he was sandbagging, 50x more throughput per megawatt, 35x cheaper tokens, and the real kicker is that agentic AI coding now eats up half of all AI queries, exactly the workload this beast was built for.

But hardware fireworks aren't the only story today: the Pentagon is threatening to blacklist Anthropic over its refusal to enable mass surveillance and autonomous weapons, a move that could redraw the ethical battle lines for every major AI lab.

Meanwhile, China turned its biggest TV event of the year into a humanoid robotics flex - kung fu robots doing backflips in front of hundreds of millions of viewers - while OpenClaw's creator just defected to OpenAI to build the AI agents everyone keeps promising but nobody has delivered yet. And if you think degrees still protect your career, a former Google exec has some uncomfortable math for you.

Grab your coffee and scroll - today's issue hits different.

All the best,

Former Google AI leader Jad Tarifi warns that long degrees like law, medicine, and even PhDs may become outdated before students graduate, as AI rapidly reaches PhD-level performance. With 70% of AI PhDs now heading into private sector jobs (up from 20% two decades ago) and some landing high six-figure offers, the traditional education-to-career pipeline is being disrupted fast. Tarifi argues Gen Z should prioritize human connection, emotional intelligence, and niche AI applications over collecting credentials, because in a world where GPT-5 rivals experts, adaptability may matter more than diplomas.

This clip is 100% AI generated. And its mindblowing good.

Terry Tao - Machine assistance and the future of research mathematics

The age of cheap AI inference just got a massive upgrade. NVIDIA's latest Blackwell Ultra platform, the GB300 NVL72, is delivering up to 50x higher throughput per megawatt and 35x lower cost per token compared to the previous Hopper generation, according to fresh benchmark data from SemiAnalysis's InferenceX v2 suite. That's not marketing fluff but independent testing across nearly 1,000 GPUs confirms Jensen Huang actually underpromised at GTC 2024 when he claimed 30x gains. The real-world numbers landed even higher.

So why does this matter right now? AI coding assistants and agentic workflows have exploded from 11% to roughly 50% of all AI queries in just one year, according to OpenRouter. These workloads are latency-hungry and context-heavy, exactly where Blackwell Ultra shines. Microsoft, CoreWeave, and Oracle are already deploying GB300 NVL72 racks at scale, with systems priced around $3 million per rack. The secret sauce isn't just silicon: NVIDIA's codesigned software stack, TensorRT-LLM, Dynamo, and NVLink's 72-GPU fabric, turns raw hardware power into real-world token economics that make previously impossible applications commercially viable.

With Vera Rubin promising another 10x leap on the horizon, the question becomes: how fast will the cost of intelligence keep falling - and what new applications will that unlock?

Why it matters: Blackwell Ultra's 35x cost reduction at low latency fundamentally changes which agentic AI applications are commercially viable at scale. As inference spending overtakes training budgets, the economics of token generation are becoming the defining competitive advantage in AI infrastructure.

Sources:
🔗 https://blogs.nvidia.com/blog/data-blackwell-ultra-performance-lower-cost-agentic-ai/

Imagine this. You open your phone to an alert. It says, “you spent $236,000,000 more this month than you did last month.”

If you were the top bidder at Sotheby’s fall auctions, it could be reality.

Sounds crazy, right? But when the ultra-wealthy spend staggering amounts on blue-chip art, it’s not just for decoration.

The scarcity of these treasured artworks has helped drive their prices, in exceptional cases, to thin-air heights, without moving in lockstep with other asset classes.

The contemporary and post war segments have even outpaced the S&P 500 overall since 1995.*

Now, over 70,000 people have invested $1.2 billion+ across 500 iconic artworks featuring Banksy, Basquiat, Picasso, and more.

How? You don’t need Medici money to invest in multimillion dollar artworks with Masterworks.

Thousands of members have gotten annualized net returns like 14.6%, 17.6%, and 17.8% from 26 sales to date.

Explore offerings

_{*Based on Masterworks data. Past performance is not indicative of future returns. Important Reg A disclosures: masterworks.com/cd}

Kung fu-fighting robots wielding nunchucks alongside children on live television, that was China's Monday night prime time. The 2026 CCTV Spring Festival Gala, watched by hundreds of millions and comparable to the Super Bowl in cultural weight, turned into a full-blown showcase for the country's humanoid robotics ambitions. Four startups, Unitree Robotics, Galbot, Noetix, and MagicLab, dominated the first three sketches of the evening, performing everything from "drunken boxing" martial arts to comedy skits and synchronized dance routines.

The progress in just twelve months is remarkable. Last year, Unitree's robots twirled handkerchiefs. This year, they executed 3-meter aerial flips, high-speed formation changes at 4 meters per second, and autonomous fault recovery, getting back up after falling down. Behind the spectacle lies serious industrial policy: China shipped 90% of the roughly 13,000 humanoid robots sold globally last year, and Morgan Stanley projects domestic sales will more than double to 28,000 units in 2026. President Xi Jinping has personally met five robotics founders in the past year, and both Unitree and AgiBot are preparing IPOs. Even Elon Musk acknowledged the competition, calling Chinese companies "ass-kicker next level."

China is turning its biggest cultural event into a launchpad for robotics industrial policy, signaling that humanoid robots are no longer R&D curiosities but a strategic priority backed by government orders, investor capital, and massive public attention. The gap between stage tricks and factory deployment is closing fast.

This is the easiest way for a busy person wanting to learn AI in as little time as possible:

Sign up to start learning.