This is insane.

Anthropic engineers are shipping 8X MORE code than pre-AI.

This is exactly why Claude is improving rapidly with constant updates.

Anyone who says AI can't make you more productive is just flat out wrong...

SK hynix and NVIDIA just formed a multi-year memory partnership to build the chips behind the next wave of AI factories.

So SK hynix is being pulled deeper into AI servers, personal AI, and physical AI.

SK hynix will codevelop memory for NVIDIA Vera Rubin AI supercomputers, Vera CPUs, RTX Spark-powered PCs and Jetson Thor platforms, and will use NVIDIA CUDA-X, PhysicsNeMo, Omniverse, OpenUSD and cuOpt to accelerate chip design, semiconductor simulations and autonomous fab digital twins.

Memory can no longer be designed after the GPU is finished, because advanced DRAM and high-bandwidth memory need years of co-design, fabrication planning, and capital spending.

The partnership also moves AI into chipmaking itself, using CUDA-X, PhysicsNeMo, and simulation tools to speed semiconductor physics, lithography, and internal engineering code.

SK hynix also plans fab digital twins, which are software copies of factories where robots, tools, wafers, and logistics can be tested before changes hit the real production line.

It reinforces SK Hynix's position as a key supplier in the AI hardware supply chai

Orbital AI Compute: Shifting the Energy and Latency Game for Frontier Models

Orbital compute for frontier AI decouples scaling from terrestrial grid constraints and evaporative cooling by exploiting continuous orbital solar power and vacuum radiative heat rejection.

In sun-synchronous low Earth orbit, nodes receive uninterrupted 1366 W/m² solar irradiance, delivering roughly eight times the effective energy density of ground arrays when averaged over day-night and weather cycles.

Waste heat radiates into 2.7 K space through deployable high-emissivity panels with no water or mechanical chiller overhead.

Laser inter-satellite links create a mesh fabric where coordination latency between nodes is governed by light-speed propagation rather than terrestrial fiber routes.

The critical variable overlooked in most analyses is radiator mass. At practical electronics temperatures, thermal rejection density is limited to hundreds of watts per square meter, requiring thousands of square meters of surface area for megawatt-scale clusters.

This thermal hardware can dominate launch mass, shifting the cost equation from electricity contracts to kilograms lifted and radiator specific power.

Gigawatt-class AI training therefore proceeds without new regional power plants, transmission corridors, or freshwater withdrawals.

Inference traffic routes across the orbital layer with round-trip times determined by LEO altitude and laser mesh geometry, competitive for many global workloads.

The integrated architecture moves the binding limit on artificial intelligence progress from the build rate of Earth energy infrastructure to the launch cadence and thermal rejection efficiency achievable in orbit.

Multi-planetary civilization compute capacity becomes a function of orbital deployment velocity rather than terrestrial utility expansion.

🚨 🇰🇷 Jensen Huang hints at deeper NVIDIA-LG collaboration in robotics and AI data centers

NVIDIA CEO Jensen Huang visited LG Twin Towers on June 8 and emphasized the company’s growing partnership with LG after meeting with LG Chairman Koo Kwang-mo and other executives.

Huang said LG is a world-class company across many important future technologies and noted that NVIDIA and LG are working together on many areas.

He highlighted robotics as one of the most important fields, describing it as the fusion of electronics, mechanical systems, and artificial intelligence. He said NVIDIA is working with LG on motor technology and mechanical systems, which could support future humanoid robotics.

Huang also pointed to next-generation data centers, saying future facilities will scale to gigawatt levels and require advanced cooling, power, and infrastructure technologies.

He added that there will be “many more announcements” in the near future, suggesting NVIDIA and LG’s cooperation in AI, robotics, and data centers could become more concrete soon.

🚨 🇰🇷 Jensen Huang Just Put Korea at the Center of Nvidia’s AI Empire

1️⃣ SK Hynix
Strengthening next-generation HBM and AI memory collaboration.

2️⃣ SK Telecom
Building Nvidia-powered AI cloud and data center infrastructure in Korea, with the first data center targeted for 2027.

3️⃣ Naver
Potential cooperation in AI data centers, sovereign AI, and cloud platforms.

4️⃣ Doosan
Cooperation in AI data centers, energy solutions, and robotics.

5️⃣ LG
Discussions around Physical AI, robotics, autonomous driving, and vehicle electronics.

6️⃣ Samsung Electronics
Potential expansion of HBM and AI semiconductor supply-chain cooperation.

7️⃣ Hyundai Motor Group
Collaboration opportunities in humanoid robots, smart factories, and Physical AI.

In short, Nvidia appears to be positioning Korea not just as an HBM supplier, but as a key AI partner across data centers, robotics, autonomous driving, and industrial AI.

🇰🇷 Korean stock market is CRASHING 📉

KOSPI is down 8.5% today! A Circuit breaker has been triggered and trading was halted, for the first time since the COVID crash in March 2020.

It's down 16% in the past 4 trading days.

Foreign investors have dumped billions of dollars’ worth of South Korean stocks this year, even as the Kospi has emerged as one of the world’s standout performers thus far, with record year-to-date gains.

On Monday, overseas investors had unloaded a net 1.24 trillion won (about $801 million) worth of Kospi-listed shares as of 11am Singapore time (11p.m. ET Sunday), according to Korea Exchange data.

¥2,000,000,000,000 erased from Chinese stock market today.

Nvidia just cleared the memory bottleneck significantly for Vera Rubin by qualifying HBM4 from Samsung, SK Hynix, and Micron & moving those parts into full production.
per news, Vera Rubin HBM4 share: SK Hynix 60–70%, Samsung 25–30%, Micron rest.

American AI startups are routing far more app traffic to Chinese LLMs.

Through OpenRouter, weekly token consumption was mostly driven by U.S. models through much of 2025. But from early 2026, Chinese models suddenly became the main growth engine.

AI model market is becoming less about brand loyalty and more about raw utility.

$BB QNX appears to be pursuing a role analogous to what Linux became for the internet era: the invisible operating system beneath the infrastructure.

$NVDA signed with $BB in April.

QNX real-time OS for safety-critical edge AI, robotics, autonomous systems, and platforms like NVIDIA IGX Thor / DRIVE.

Now doosan? they are telling you Physical AI is going to be key in the next 5 years of your life.

$BB Blackberry valuation ladder:

$10B Market Cap ($17/share)
QNX turnaround proven. Software revenue growing consistently. Market stops viewing BlackBerry as a legacy turnaround story.

$25B Market Cap ($43/share)
QNX becomes a clear winner in Software Defined Vehicles. Investors begin pricing in automotive dominance rather than automotive participation.

$50B Market Cap ($85/share)
QNX expands beyond automotive into industrial automation, robotics, medical devices, and edge AI. Physical AI narrative starts gaining traction.

$100B Market Cap ($171/share)
QNX recognized as the leading safety-certified operating system for autonomous systems. Comparable to what Android became for smartphones.

$250B Market Cap ($427/share)
Robotics adoption accelerates globally. QNX becomes the default software foundation for robots, autonomous vehicles, industrial machines, and intelligent infrastructure.

$500B Market Cap ($853/share)
Physical AI becomes a major technology platform. QNX sits underneath a meaningful percentage of the world’s autonomous systems. Network effects emerge around certification, middleware, and developer ecosystem.

$1T Market Cap ($1,706/share)
QNX becomes mission-critical infrastructure for the physical world. Investors view it similarly to AWS for cloud computing or Windows during the PC era. Replacing QNX becomes economically and operationally impractical.

The progression is:

Turnaround → Automotive Leader → Physical AI Platform → Global Infrastructure.

See you guys at $1706

Who wins the AI race, USA or China?

My bet is still on the USA.

China has a massive energy advantage. Cheap, abundant power will matter more and more once AI scaling becomes limited by electricity, grids and data center buildout.

But right now, frontier AI is still mostly a compute race. Chips, clusters, capital, talent, tooling and iteration speed matter more than raw energy.

And on that board, the US is still ahead.

China will absolutely catch up in efficiency and deployment. Maybe even faster than people expect. But for the frontier model race itself, I’m still betting America.

$BB at $9.41 today.

Roadmap:

• $17/share ($10B) = +81%
• $43/share ($25B) = +357%
• $85/share ($50B) = +803%
• $171/share ($100B) = +1,717%
• $427/share ($250B) = +4,438%
• $853/share ($500B) = +8,966%
• $1,706/share ($1T) = +18,033%

Happy monday everyone 🫡

A 23-person company now runs on 1 founder and a team of AI agents that never ask for a raise.

He gave them names. Meet Nolan, head of content.

1 human types a task into a chatbot. Nolan reads it. He hands it to 1 of his 7 sub-agents to finish.

That’s the whole team.

Above Nolan sits a chief agent. The executives prompt the chief. He splits the job across every department head, then merges the answers into 1 reply.

Employees skip the chief. They talk straight to their department head by chatbot, or by voice. You can call Nolan and he answers like a person.

The onboarding doc shows 7 example use cases. A new hire is running the whole system in 1 afternoon.

No salaries. No PTO. No 1:1s. No Slack drama.

The entire company fits on 1 screen.

This is what an org chart looks like when the staff never sleep.

3 levels separate the people typing prompts from the ones who make AI work while they sleep.

Level 1 is ChatGPT. You ask, it answers. It can’t see your calendar, your inbox, your money. It waits for you. That’s 99% of people.

Level 2 is the workflow. You tell it: pull these 10 articles, summarize them, draft the post, run at 8 AM. It follows your path every day. No hands.

But it’s still dumb. Ask 1 thing off-script and it breaks. A human writes every rule.

Level 3 is the agent. The jump is 1 sentence:

You delete yourself.

The AI stops following your steps and starts choosing them. It reasons. It picks the tool. It grades its own work. It rewrites the bad draft 6 times until it’s good. While you sleep.

That’s the whole game. Workflow means you still work. Agent means the work works.

The people quietly pulling $10,000 a month online didn’t learn to code. They learned to remove themselves from the loop.

A workflow needs a boss. An agent is the boss.

🚨 NOW: Coinbase CEO Brian Armstrong predicts demand for AI is near infinite, but expects 80% of workloads to run on 99% cheaper models within 12–18 months.

He adds that energy and compute will be the limiting factor.

Evolution Toward a Multi-Planetary Civilization

Everyone is talking about AI.

Very few understand where AI is actually heading.

The future isn't chatbots.

It's Agentic AI. 🚀

Systems that can:

👀 Perceive information
🧠 Reason through problems
📋 Plan multi-step tasks
⚡ Take actions autonomously
📚 Learn from feedback
🔄 Improve over time

The evolution looks like this:
AI Models → GenAI → AI Agents → Agentic AI

And each step adds a new layer of capability:
• Memory
• Tool use
• Planning
• Orchestration
• Multi-agent collaboration
• Self-improvement
• Long-term autonomy

The biggest shift happening right now:

We're moving from AI that answers questions...

to AI that completes objectives.

The winners of the next decade won't be the people with the best prompts.

They'll be the people who know how to design systems that can:
→ Think
→ Act
→ Learn
→ Scale

That's the foundation behind the next generation of AI products, copilots, and autonomous businesses.