The proposed U.S.-Iran settlement arrived Friday in the familiar costume of modern diplomacy: nearly final, heavily sourced, immediately disputed and already moving markets.

Pakistan's prime minister, Shehbaz Sharif, said a final agreed text had been reached between Washington and Tehran, with mediators working on next steps. The emerging arrangement, as described by regional officials and U.S. briefers, is supposed to reopen the Strait of Hormuz, start a phased easing of sanctions, release frozen Iranian assets and kick the problem of Iran's enriched uranium into a 60-day technical process. That is not nothing. It is also not peace, at least not yet. It is a machine for discovering whether two governments mean the same thing when they use the same nouns.

On the American side, the word being sold hardest is performance. A senior U.S. official told reporters Iran would have to begin delivering concessions before receiving benefits. That sequencing matters because Washington is trying to turn the agreement into leverage rather than tribute. In the U.S. telling, the strait reopens, the blockade lifts, uranium removal gets engineered and money follows compliance.

Tehran is already describing a narrower object. Iranian Foreign Minister Abbas Araqchi said nuclear talks belong to a later stage, not the current interim agreement. He also said management of the Strait of Hormuz will not simply return to its prewar pattern, with Iran claiming a continuing security role over passage. That is the soft underbelly of the deal: the route that carries energy to the world may reopen without anyone agreeing who owns the threat of closing it again.

Lebanon is the other trapdoor. Iranian officials have framed the memorandum as including an end to fighting there, which would imply Israeli withdrawal from occupied areas. Israel's defense minister has said Israel will not withdraw from security zones in Lebanon, Syria or Gaza. The deal therefore contains, or is being made to contain, a promise by one side about the posture of a third. That is how agreements become grievance generators.

The market, less offended by ambiguity than civilians are, cheered the possibility of reduced risk. Brent crude fell, U.S. stocks rose and the SpaceX debut helped add a second kind of intoxication to Friday's trading. For investors, even an unsigned pathway to more oil flow is tradable relief. For sailors, Lebanese villagers, Iranian families and American service members, it is a schedule of hazards.

The Strait of Hormuz itself has already changed. AP analysis found more Gulf Arab oil appears to be reaching market with U.S. military support while Washington continues to squeeze Iran's sanctioned oil trade. A deal that reopens the waterway may not restore the old waterway. Insurers, shippers and navies have learned new habits. Iran has learned the power of threat without full closure. Gulf states have learned how quickly the business model of stability can become a convoy problem.

So the headline is not that peace has arrived. The headline is that diplomacy has found a form sturdy enough for each side to stand on while saying different things. That can still save lives. Many wars end because exhaustion learns to speak in footnotes. But this agreement, if signed, will begin as a test of enforcement, not trust. The memorandum may be a door. It is also a hallway lined with tripwires.

China's arrest of Min Zin, an American scholar and Myanmar-born political analyst, is a small story only if diplomacy is measured by banquet seating.

Beijing said Friday that Min Zin had been detained on suspicion of espionage and endangering national security. AP reported that he had traveled to Kunming, in Yunnan province, for a conference and disappeared June 3, according to an activist familiar with the case. The U.S. State Department confirmed that Min Zin, a U.S. citizen, was detained while in China and said consular officials had visited him.

The timing matters. The detention comes about a month after President Donald Trump met Xi Jinping in Beijing, a summit advertised as a reset of a relationship that keeps refusing to reset. The accusation also lands in a region where China does not draw neat lines between academic work, border politics, resource flows and national security.

Min Zin is not just a random visiting scholar. He was a student activist during Myanmar's 1988 uprising, later sought asylum in the United States, founded ISP Myanmar and has studied Chinese foreign policy, trade and Myanmar's rare earth exports to China. That profile sits exactly where Beijing's sensitivity is highest: democracy movements, border influence, minerals, think tanks and foreign citizenship.

Espionage allegations in China often operate as both legal claims and political weather systems. For Washington, the case becomes another detained-American file. For Beijing, it is a warning to the class of people who believe regional expertise can remain safely between states. For Myanmar's exiles and analysts, it is a reminder that proximity to China is not merely geographic.

The larger U.S.-China relationship is now thick with transactional calm: summits, trade management, AI competition, Taiwan choreography. Min Zin's case slices through that calm because it personalizes the risk. The reset has a human hostage problem before anyone wants to admit it does.

A federal appeals court rejected the Kennedy Center leadership's last attempt to keep President Donald Trump's name on the building Friday, leaving workers to do what courts, arts boards and presidents had turned into a national ritual: remove a nameplate.

The order followed U.S. District Judge Christopher Cooper's ruling that Trump's name had been illegally added to the Washington performing arts venue and had to be removed by Friday. AP reported that workers erected scaffolding near the sign while a crowd gathered nearby. The Kennedy Center website had already dropped the Trump branding, and staff had been instructed to use the center's statutory name, The John F. Kennedy Center for the Performing Arts.

The facts are plain; the symbolism is loud. Trump returned to office and rapidly asserted influence over the institution, replacing leadership and installing a board that named him chairman. The court found the renaming exceeded lawful authority. Cooper also blocked a planned two-year closure for major renovations.

This was never simply an argument over typography. Public buildings use names as civic memory. They also use them as possession. Trump's effort to staple his own brand to the Kennedy Center compressed several of his governing instincts into one facade: personnel control, cultural combat, institutional rebranding and impatience with process.

The court's rebuke does not depoliticize the arts. That ship sailed long ago. But it does draw a line around how culture is claimed by power. Congress created and named the center. A president may influence it, bully it, fund it, attend it or neglect it. He does not get to treat it like a hotel lobby.

The scaffolding was the image of the day in Washington because it turned abstraction into labor. Somebody had to climb up and undo the boast.

SpaceX did not merely go public Friday. It entered the machinery that makes public markets feel inevitable.

Shares jumped more than 19 percent in their first session, closing just below $161 after opening around midday at $150, according to AP. That put SpaceX at roughly a $2.1 trillion market value, larger than Tesla and, by AP's comparison, bigger than Exxon Mobil, Bank of America and Coca-Cola combined. Forbes estimated Elon Musk's net worth at about $1.1 trillion after the debut, making him the world's first trillionaire.

The company sold investors a familiar but unusually cinematic proposition: ignore current losses, pay for the future. SpaceX is asking markets to fund satellites, orbital data centers, artificial intelligence and the old Mars promise, refreshed for a new liquidity cycle. Musk rang the bell remotely from Starbase and spoke again of making life multiplanetary. Wall Street heard, among other things, index inclusion.

That detail deserves more attention than the ceremonial bell. AP reported that Nasdaq revised its rules to allow SpaceX to enter funds tied to its indexes in 15 days. The practical effect is that ordinary retirement savers and passive investors may own the stock quickly, not because they chose SpaceX as a speculative bet but because the benchmark did.

There is nothing inherently improper about that. Index funds are the plumbing of modern wealth. But plumbing is not neutral when it starts carrying rocket fuel. A company whose value depends on government contracts, launch dominance, broadband satellites, AI infrastructure and charismatic risk now becomes part of the default savings architecture. The market's wager becomes the public's ambient exposure.

The broader tape helped. U.S. stocks rose Friday as oil prices eased on hopes for a U.S.-Iran deal. The S&P 500 gained 0.5 percent, the Dow added 0.7 percent and the Nasdaq rose 0.3 percent. The artificial intelligence trade, battered earlier in the week, got a glittering rescue narrative: here was a company that could be space, defense, data centers and AI all at once.

That is the bull case and the warning label. SpaceX may be a great company. It may also be a great excuse for investors to stop distinguishing between categories. The public market has now priced the future as if orbit is a software margin waiting to happen. Victor Ledger's rule remains unfashionable but useful: when a company promises civilization, read the cash-flow statement twice.

The useful thing about a shopping cart was that it waited for you. The new agent economy would prefer to keep moving.

Visa said this week it has embedded its payment network inside ChatGPT, allowing AI agents to shop and complete transactions on behalf of users. AP reported that the integration could let an agent not only recommend products but buy them at potentially any merchant that accepts Visa. OpenAI supplies the agent technology; Visa supplies authorization and fraud monitoring.

This is the boundary crossing that matters. Chatbots have been annoying, impressive, wrong, helpful and occasionally uncanny. But when they can spend, their mistakes acquire receipts. A hallucinated product, a misunderstood preference or an overly literal instruction becomes less like bad search and more like a disputed charge.

Visa and OpenAI will describe the system in the language of guardrails, permissions and monitoring. They should. Payments cannot scale on vibes. Still, consumer trust will be won in the after-sales experience, not the launch demo. Can users see why an agent chose a vendor? Can they limit categories, amounts and frequency? Can they reverse a purchase without being trapped between a card network, a merchant and an AI interface that says it followed instructions?

The move also reframes merchant power. OpenAI's earlier Instant Checkout effort struggled partly because merchants resisted the fees and the funnel. Visa brings a far broader acceptance layer. If this version works, the agent is not a feature on a store. It becomes the front door to many stores.

That could be magnificent for replenishing diapers and grim for everything requiring taste, restraint or doubt. The machine will not just answer the question. It will ask for a card.

Anthropic is now paying to study the disruption it is helping accelerate.

The company announced an initial $200 million commitment to research AI's impact on jobs and the economy, along with policy experiments it considers promising. AP also reported a $150 million national fellowship program meant to place early-career workers into community-serving organizations, and a separate Claude Corps plan to embed 1,000 fellows trained on Claude at nonprofits and other groups for a year.

The humane reading is that this is what responsible firms should do. Anthropic's CEO, Dario Amodei, has argued that AI could create larger and longer labor disruptions than earlier technologies and has floated economic support, including universal basic income, as part of the policy conversation. The company is putting money behind the social problem instead of pretending the market will automatically make everyone whole.

The harder reading is that private AI labs are becoming both accelerants and paramedics. They build the systems, shape the risk, fund the studies, train the adapters and propose the safety net. That is not hypocrisy so much as governance by whoever got rich first.

The fellowships may do useful work. Many nonprofits, schools and local governments genuinely need help using AI without being eaten by vendor jargon. Anthropic's program, managed with CodePath, says it will not require a narrow credential path and will place fellows in organizations that serve broad communities. That is better than another demo day for executives.

But the country's AI labor policy cannot be outsourced to corporate conscience. A reader recently wrote to us arguing that oil may be cheap because the economy is no longer fundamentally oil-driven, that everything is now about AI. Friday's issue is an argument in favor of that suspicion. Oil still moves prices. But AI now moves jobs, payments, IPO calendars, nonprofit staffing and the excuses politicians use to preempt state law.

The future of work is no longer a white paper. It is a budget line written by the companies making the weather.

The official World Cup machine wants scale: stadium lights, celebrity arithmetic, anthem-by-committee uplift. Bosnia found something sharper, funnier and harder to manufacture: a song with a past.

Dubioza Kolektiv's 2011 track "USA" began as a satire of immigration longing and disillusionment with the American Dream. Its rewritten incarnation, "I Am From Bosnia, Take Me to America," has become a viral soundtrack for Bosnia-Herzegovina's World Cup run. AP reported that the new video drew nearly 2 million views in less than three weeks, on top of the original song's long afterlife.

The trick is not subtle, which is why it works. The lyric that once sounded like a plea to leave becomes, in a World Cup hosted partly in North America, a chant of arrival. The joke turns around and salutes itself. A country with a difficult history and only its second World Cup appearance gets to sing desire without humiliation.

This is what sports pageantry does best when it escapes the marketing department. It gives ordinary people permission to convert contradiction into rhythm. Bosnia's fans are not rejecting the American Dream exactly. They are using it as a prop, a destination, a punchline and a field assignment. The song carries the melancholy of migration and the cheap thrill of a terrace chorus at the same time.

A reader asked us recently whether people are even watching FIFA anymore, whether sports now feel passe. The answer is yes, people are watching, but not always in the old way. The matches are content. The songs, memes, watch parties and diaspora rituals are the deeper broadcast. Geopolitics enters wearing a scarf. National identity arrives with a hook.

The best anthem is rarely the one chosen to represent everyone. It is the one that gives a specific crowd the feeling that it has smuggled its private joke into the global feed. Bosnia's song does exactly that. It makes the World Cup feel less like a property and more like a place people can still trespass emotionally.

Elon Musk becoming a trillionaire will be described in the passive voice, as if a mountain formed or a storm system gathered strength over the Atlantic. This is the first mistake.

A trillion-dollar fortune is not weather. It is law, subsidy, defense procurement, monetary policy, patent protection, index design, labor leverage, investor credulity and a public that has been trained to call concentrated power innovation when the logo is sufficiently futuristic.

SpaceX may deserve admiration. Rockets are difficult. Reusable rockets are a civilizational achievement. But the market's Friday ecstasy did not merely reward engineering. It converted a founder's myth into a default holding for passive investors and pension accounts. The man became a trillionaire not outside society, but through society's deepest pipes.

This is why the usual argument between envy and merit is so boring. The question is not whether Musk worked hard. The question is why any democratic economy should be so casual about fortunes that begin to resemble private ministries. A trillionaire can shape communications infrastructure, defense priorities, AI capacity, transport fantasies, political speech and labor markets before breakfast.

The public is invited to applaud because the rockets are beautiful. They are. But beauty is not accountability. Awe is not governance. The number is not a cloud. It is a constitutional problem wearing a spacesuit.

The memorandum of understanding is the perfect instrument for a world that wants peace without the shame of trust.

A treaty is too solemn. A ceasefire is too measurable. A surrender is too honest. The MOU, by contrast, is fog with formatting. It lets every side announce momentum while privately preserving its favorite betrayal. Washington can say Iran agreed. Tehran can say it agreed to something else. Israel can deny the implication. Markets can buy the rumor and sell the refinery.

This is not an argument against the U.S.-Iran document. If the paper reopens Hormuz, slows killing and buys time, then by all means print it, sign it digitally, laminate it and carry it through customs. People live inside pauses that pundits dismiss.

But we should stop treating ambiguity as wisdom merely because diplomats produced it. The modern conflict-management state has grown addicted to interim arrangements that postpone the moral question. Who controls the strait? Who pays for the dead? Who disarms? Who leaves Lebanon? Who verifies uranium? Who is punished if the answer changes after the cameras leave?

The MOU is useful precisely because it does not answer everything. It is dangerous for the same reason. Peace is not the absence of explosions on a trading screen. It is the day after the document discovers whether it has teeth.