1. SpaceX Goes Public (Space)
SpaceX priced at $135, opened at $150, traded above $176, and closed its first session around $161. The company raised about $75 billion, making it the largest IPO on record. 1
The second day pushed the move further. By Monday, SpaceX closed above $192, putting its market value near $2.6 trillion. That made it one of the most valuable public companies in the US almost immediately.
The public float is still tiny. Only about 4.3% of shares are freely tradable, with large insider and early-investor lockups coming later. 2
That matters because investors are not buying a normal space company. They are buying Starlink cash flow, launch dominance, Starship optionality, and an AI infrastructure bet under one ticker.
Why it matters
The largest IPO ever turned a private frontier company into a public market anchor. SpaceX is now a stock, an index problem, a retail obsession, and a liquid currency for Musk's larger AI and space ambitions.
Reality check
A first-day pop is not proof. The float is small, the valuation is huge, and future unlocks will test demand. Public shareholders get exposure to the whole machine, but not much control over it.
2. The Model the Government Pulled (AI)
On June 12, Anthropic received a US export-control directive ordering it to suspend access to Fable 5 and Mythos 5 for any foreign national, even foreign-national Anthropic employees. Anthropic said the only practical way to comply was to disable both models for everyone. 3
The government did not give a detailed public explanation. Anthropic said officials appeared to be reacting to a narrow jailbreak claim, not a universal break of the model's safeguards. The company said the demonstrated vulnerabilities were minor and available from other models too.
The backlash came quickly. More than 100 cybersecurity executives and experts urged the administration to ease the restrictions, arguing that pulling the models could hurt US cyber defense more than it helps. 4
Why it matters
Frontier AI crossed into export-control logic. A cloud model was treated less like software and more like strategic capability. That changes how customers, allies, employees, and rivals think about depending on American AI.
Reality check
The public still does not know how serious the jailbreak was. Anthropic has every reason to downplay it, and the government has not shown its work. The order may be a necessary warning shot, a blunt overreaction, or both.
3. Hormuz Reopens, Maybe (Crypto / Geopolitics)
Trump announced an interim agreement with Iran to reopen the Strait of Hormuz and move toward ending the war. Markets reacted immediately: US stocks jumped, oil fell, and risk appetite came back. 5
The deal is not clean yet. A formal signing is expected later, and shipping groups are still warning that safe passage is not fully normalized. Iran's proposed fees and control claims remain unresolved. 6
Crypto moved with the relief trade. Bitcoin recovered from the low $60,000s as traders priced lower energy risk, lower inflation pressure, and less chance of a wider conflict.
The same G7 week brought a second peace signal: Trump said he had spoken with Zelenskyy and Putin and believed there was room to move on Ukraine. That does not end either war, but it gave markets a cleaner story: lower energy risk, lower inflation pressure, and less geopolitical tail risk. 7
Why it matters
The market story is sustained relief. If Hormuz reopens and Ukraine diplomacy improves, the pressure comes off oil, inflation, rates, and risk assets at the same time. That is why crypto bounced with stocks instead of trading like a separate world.
Reality check
This is still fragile. Announcements are not ships moving safely through the strait, and Ukraine remains an active war. The rally can reverse if the Iran deal slips, shipping risk stays high, or diplomacy turns back into another headline cycle.
4. Europe Buys a Humanoid Contender (Robotics)
Germany's Neura Robotics raised $1.4 billion at about a $7 billion valuation. Backers include Amazon, Nvidia, Qualcomm, Tether, Bosch, Schaeffler, and the European Investment Bank. 8
Neura wants to build industrial AI and humanoid robots from Europe. Its near-term target is to expand humanoid production from 6,000 units this year to tens of thousands next year, with a longer-term goal of millions of robots and robotic arms by 2030.
The company is also building Neuraverse, a training platform where robots learn tasks from humans in motion-capture suits. The pitch is simple: Europe cannot own physical AI unless it owns the robots, the training data, and the manufacturing path.
Why it matters
The humanoid race is no longer just US versus China. Europe is trying to turn industrial depth into a robotics stack before the category hardens around American software and Chinese hardware.
Reality check
Money is not manufacturing. Scaling from thousands to tens of thousands is the hard part, and humanoids still have to prove useful work at acceptable cost. Europe has strong suppliers, but it has not been great at scaling frontier hardware companies.