BREAKING: US intelligence assesses that Iran's new Supreme Leader Mojtaba Khamenei is far more interested in a nuclear weapon than his father and is looking for ways to pursue an advanced thermonuclear warhead that could be miniaturized and placed on a long-range missile, per US officials to NYT.
US intelligence adds that Iranian officials who previously argued only a nuclear weapon could prevent an Israeli or American attack believe their position has been significantly strengthened.
JUST IN: 🇨🇳🇺🇸🇮🇷 China responded to Trump’s threats over the support for Iran:
“As a comprehensive strategic partner, China strongly supports Iran in safeguarding its sovereignty, security, and national dignity”
JUST IN: 🇺🇸🇮🇷 US intelligence finds that Iran's new Supreme Leader Mojtaba Khamene is "far more interested" in pursuing a nuclear weapon than his father was, NYT reports.
🚨 WE ARE ABOUT TO WITNESS THE BIGGEST WAR IN HISTORY FOR THE SURVIVAL OF PETRODOLLAR
Massive U.S. bomber formations, continuous surveillance flights, and heavy air movements are now saturating the skies near Iran. This is no longer posturing.
This is the opening chapter of the biggest geopolitical-financial war of our lifetime, the one that will decide whether the petrodollar survives… or dies.
If the U.S. fails to completely dismantle the IRGC’s control inside Iran, the consequences are not gradual. They are terminal‼️
Iran has already laid down the new law of energy: Every single barrel of oil that passes through the Strait of Hormuz, the artery carrying ~20% of the world’s crude, must be settled in Chinese yuan. No dollars. Full stop.
The moment that demand sticks, the 50-year monopoly that forced the world to recycle oil revenues into U.S. Treasuries collapses overnight. The petrodollar doesn’t weaken. It dies. And the yuan (backed by real oil flow) becomes the new settlement currency of the energy world.
Watch the bombers. Watch the surveillance orbits. Watch Hormuz.
Because if this war ends with the IRGC still standing, the dollar’s exorbitant privilege ends with it and a multipolar monetary order ruled by Beijing’s currency and decentralized rails takes its place.
The markets haven’t priced this yet. They will.
The faith of the petrodollar is now being decided in the skies over the Gulf.
Jensen Huang just exposed the one thing that could kill NVIDIA's empire, and he's begging the industry to stop it (Save this).
NVIDIA published a letter explaining why open weight AI models matter, and buried inside is the real reason Jensen cares so much about this fight.
The letter opens by comparing this moment to the 1980s open source software movement, arguing that just as open code became the foundation of the modern internet, open weight models will become the foundation of American AI leadership.
It argues America's AI leadership won't be decided by one frontier model, but by whether the country builds an open ecosystem that spreads into every industry, from factories to hospitals to small businesses.
Open weights let any startup or institution match the right model to the right job without paying frontier prices for every task, which keeps AI spending distributed instead of locked up in a few giant labs.
NVIDIA also argues open weights give customers real control, since companies that build on open models don't risk getting locked into a single provider and can deploy models wherever their business actually needs them to run.
Then the letter openly admits the downside, acknowledging that once weights are released, they're out of the original developer's control and modified versions are hard to trace or reverse.
But NVIDIA's answer to that risk is not to restrict open models, it argues that in a world where attackers already use advanced AI, defenders need equally capable open tools to detect, simulate, and respond to emerging threats.
Now here's the part that actually explains Jensen's real motivation.
Chinese open source models like DeepSeek and Kimi are still trained on NVIDIA GPUs and released in forms optimized for NVIDIA hardware which means even AI progress NVIDIA doesn't fully control still ends up reinforcing NVIDIA's dominance.
Nobody is running that inference on Amazon's Trainium chips instead, because the models simply aren't built for that hardware.
That dynamic protects NVIDIA's grip on the inference market and helps the neocloud companies built entirely on its GPUs, while also making it harder for hyperscalers to justify switching to their own custom ASICs.
If China ever built real independence from NVIDIA hardware, that protection disappears entirely, NVIDIA's dominance weakens, and rivals like TPUs and Trainium suddenly look a lot more competitive by comparison.
So this isn't really about ideology or generosity toward the open-source community, it's NVIDIA defending the one structural fact that keeps every AI lab, American or Chinese, open or closed, still needing to buy its chips to actually run any of it.
That's exactly why this is bullish for neoclouds, open weight AI expands the number of companies deploying models, and most of them will rent NVIDIA powered compute from providers like Nebius and CoreWeave instead of building their own infrastructure.
Make sure to follow @MelvinInvests for more AI infrastructure and NeoCloud insights and check out the link below for more details.
🤩 A 124B model activating only 5.1B parameters while approaching its 1T sibling’s performance.
It's ~ as good as Sonnet-4.6 to run locally.
The mystery model that appeared before its announcement is official.
@TheInclusionAI has officially released Ling-3.0-flash, a hybrid-reasoning MoE built for production-scale AI agents.
Stats:
🧠 124B total parameters
⚡ Just 5.1B active per token
📚 256K native context
🔭 Designed to scale toward 1M context
InclusionAI says with roughly 1/8 the total parameters and 1/12 the active parameters, Ling-3.0-flash matches or beats its 1-trillion-parameter flagship on most of the benchmarks shown.
What is different with Ling-3?
🧬 Hybrid linear attention
Ling stacks:
🔹 5 KDA layers for efficient long-range memory
🔹 1 MLA layer for precise attention
The goal is to preserve long-context agent performance without paying full-attention costs across every layer.
InclusionAI demonstrated Ling-3.0-flash by 🏙️ building and rendering a complete 3D city through Blender MCP, 🔬 coordinating a five-agent scientific research team 📊 editing Word proposals and validating Excel financials 🎨 generating full design systems without external images, 🎹 building an interactive browser synthesizer with Web Audio 📱 adapting one source into X, Instagram and LinkedIn content and 📅 reading messages, finding availability and updating calendars through OpenClaw
It is designed to receive a goal, coordinate tools and close the workflow.
You can try it now
🆓 Live on OpenRouter
📅 Free through August 3, 2026
🔌 Model ID:
inclusionai/ling-3.0-flash
The API is live, but InclusionAI has not yet confirmed downloadable weights, a license or a consumer-local deployment path.
Localmaxxers can test its agent behavior today but cannot yet load it into LM Studio or llama.cpp.