The U.S.-Iran deal has finally acquired the one thing missing from the week of choreography around it: words that can be argued with.

The draft memorandum made public through the Associated Press sets out an immediate and permanent halt to military operations, including in Lebanon, a pledge by both sides not to use force against each other, and a sixty-day window to negotiate a final agreement. That is the headline claim. The fine print is where the power sits.

The document moves first on what the world economy could feel fastest. The United States would begin removing its naval blockade immediately and finish within thirty days. Iran would use its best efforts to restore safe commercial passage through the Strait of Hormuz, toll-free for sixty days, with traffic reinstated within thirty days as technical, military and demining work proceeds. Washington would issue waivers for Iranian crude oil, petroleum products and associated services, including banking, insurance and transport, upon signing.

The harder parts are pushed into the sixty-day room. Iran reaffirms that it will not seek nuclear weapons and agrees that the disposition of enriched material will be resolved through a mutually agreed mechanism, with downblending on Iranian soil under IAEA supervision as the minimum method. Broader sanctions termination is tied to the final deal. Frozen assets are to become usable upon implementation, not simply upon ceremony. A monitoring mechanism is promised. A final agreement would need endorsement through a binding U.N. Security Council resolution.

That architecture is not trivial. It is also not peace in the way politicians use the word when they want the public to stop reading. The memo front-loads relief to the parts of the crisis that were punishing shipping lanes, oil markets and allied politics. It back-loads the dispute over enrichment, sanctions sequencing and regional force. The wager is that commerce can create momentum where trust cannot.

The weakest line may be the one that looks practical. Iran would provide toll-free passage through Hormuz for sixty days, while future administration of the strait would be discussed with Oman and other Gulf states. Iranian officials, according to the Guardian, are already signaling that Tehran believes the strait will not simply return to prewar conditions and that fees may come later. If that becomes the new dispute, the deal will have converted a blockade crisis into a pricing argument with naval consequences.

At the G7 in France, allied leaders welcomed the tentative deal and praised President Trump’s role, according to AP. That praise bought something: a joint statement offering unwavering support for Ukraine and calling for more air defense, interceptors and long-range capabilities. It also bought time for a president who prefers theater to committee language to remain inside the room long enough for Europe to get a sentence it needed.

This is what diplomacy looks like when it has to satisfy three audiences at once. Markets want the strait open. Gulf states want shipping predictable. Washington and Tehran want to claim leverage before concessions harden into blame. The public gets a memo with enough substance to matter and enough ambiguity to keep every capital awake.

The old fog around the deal has lifted. What remains is not comfort. It is a clock, ticking toward the place where the ceremonial language runs out and the unresolved clauses start charging rent.

The most useful thing at the G7 may not have been the communique. It may have been the seating chart.

After months in which U.S. allies were angered by the Iran war and uncertain about Washington’s appetite for Ukraine, the summit in France ended with a public performance of alignment. AP reported that France, Italy, Germany and Britain, all previously critical of Trump’s decision to launch the Iran war without allied consultation, welcomed the tentative U.S.-Iran deal and credited his leadership. In return, Trump joined leaders in offering unwavering support for Ukraine and calling for more air defense systems, interceptors and long-range capabilities.

This is transactional diplomacy in its least embarrassing form. Allies flattered the ego attached to the policy they needed stabilized, then used the available goodwill to pin down language on a war where Europe carries more of the military and financial burden. Nobody should confuse that with a grand revival of the old West. It is a working arrangement, and working arrangements are underrated.

Macron’s decision to keep Trump through the full summit with a Versailles dinner was mocked as pageantry because it was pageantry. But pageantry is not the opposite of policy. Sometimes it is the room service of policy, the ornamental mechanism that keeps a difficult guest seated long enough for the staff work to happen.

The China language was less clean. G7 leaders warned against economic coercion, while Trump publicly thanked China and Russia for remaining neutral in the Iran conflict. The alliance is still doing two things at once: presenting unity on paper and absorbing improvisation at the microphone.

Still, the summit mattered because the leaders did not leave with the usual busted china. Iran got a framework. Ukraine got a sentence. Trump got applause. Europe got proof that flattery, when used with discipline, is a diplomatic instrument rather than merely a vice.

The B-52 is the kind of American machine that sounds less maintained than inherited. It entered service in 1955, is scheduled to keep flying into the middle of this century, and sits in the national imagination as proof that old steel can be made permanent if enough money and doctrine are applied.

On Monday, one of them crashed shortly after takeoff at Edwards Air Force Base in California’s Mojave Desert, killing all eight people aboard. AP reported that the aircraft was in the air only briefly before slamming into the ground about halfway down the runway. It was taking part in a test mission connected to efforts to keep the B-52 fleet flying for decades.

The aircraft was supporting a radar modernization program, according to an Air Force official cited by AP. The broader upgrade effort is meant to replace antiquated radar and add new engines, communications, avionics and weapons systems. The logic is familiar: refurbish the old platform until it can serve alongside the newest bomber, the B-21 Raider.

Investigators will examine flight controls, engines, maintenance, age and structure. Aviation experts cautioned that it is too early to name a cause, but one former F-16 pilot told AP that pilot error seemed unlikely given the test crew’s expertise and that structural issues could not be ruled out in an aircraft whose youngest airframes date to 1962.

That is the terrible poetry of the crash. The mission was about extending the life of an aircraft built before personal computers, before the moon landing, before much of the bureaucracy now defending it. Longevity can be strategic. It can also become a wager paid in other people’s bodies. The investigation will decide what failed. The age of the plane has already testified.

The Federal Reserve did not raise interest rates on Wednesday. Wall Street sold off anyway, which is a neat summary of modern monetary policy: the action matters, the punctuation matters, and the dot plot can ruin lunch.

The Fed held its benchmark range at 3.5 percent to 3.75 percent, where it has been since December. But AP reported that nearly half of policymakers now foresee at least one rate increase this year. The S&P 500 fell 1.2 percent to 7,420.10. The Dow lost 507.12 points, or 1 percent, and the Nasdaq slid 1.3 percent. Treasury yields climbed as traders repriced the possibility that the next move may not be the cut they wanted.

The Guardian reported that inflation has been pushed to 4.2 percent, the highest since 2023, after energy prices spiked during the Middle East conflict. Core inflation is milder at 2.9 percent, and unemployment remains at 4.3 percent. That mix leaves the Fed in the most irritating part of its job: headline inflation is too hot, underlying demand is not obviously broken, and markets have been trained to look for rescue before they look for discipline.

This was also Kevin Warsh’s first meeting as chair. He has signaled a desire to tighten the Fed’s public communications, which sounds procedural until traders discover that less soothing guidance means more room for fear. A central bank can hold rates and still tighten conditions if investors believe patience is no longer generous.

The Iran memorandum may bring oil relief if Hormuz traffic normalizes. But energy shocks fade through invoices, not headlines. The Fed is not paid to celebrate the first price drop. It is paid to decide whether the drop will survive the next clause, ship inspection or missile battery. Wednesday’s market learned that a hold can still have teeth.

Nvidia’s latest promise is not only that AI will change work. It is that the AI buildout will make work visible again.

In Sherman, Texas, Jensen Huang stood at a Coherent facility tied to a $2 billion partnership and argued that AI infrastructure can revive U.S. manufacturing. AP reported that the factory will produce material for lasers used to transmit data among chips, allowing them to operate with more power, speed and efficiency. Huang also told AP that the AI age requires new social norms and that people should embrace the technology’s opportunities rather than fear it.

This is the more politically durable version of the AI boom. Software eating jobs is a hard sell. Factories expanding north of Dallas are much easier to photograph. The claim moves AI out of the abstract cloud and into concrete, permits, electricians, logistics and payrolls. It gives governors something to cut ribbons around and executives something to say when asked whether automation is about replacing workers.

But infrastructure jobs and economy-wide bargaining power are different things. AI demand can support manufacturing nodes while still concentrating profits in platform owners, chip designers and utilities with control over scarce inputs. The factory floor can be real and the distribution story can still be thin.

Huang is right that a new technical system needs new norms. The missing word is ownership. Who owns the productivity gains? Who owns the local costs: power demand, water use, housing pressure, public subsidies, specialized training pipelines? A laser component plant can be part of an industrial renewal. It can also become a beautiful hinge in a supply chain whose largest rewards are booked elsewhere.

The Texas test is useful because it is material. If AI is going to ask the public for land, energy and political patience, it should have to show up in places that can count the jobs and inspect the bills.

Britain has decided that childhood should not be administered by a recommendation engine, and has chosen the blunt instrument first.

Prime Minister Keir Starmer announced that children under 16 will be banned from a range of social media apps, including Snapchat, TikTok, YouTube, Instagram, Facebook and X. The government says messaging services such as WhatsApp and Signal will not be included, and the policy is expected to take effect early next year. The U.K. says it will follow the Australian model, targeting platforms built for user-to-user posting, social interaction and algorithms.

The government is also moving against adjacent risks. AP reported that authorities plan action to prevent strangers from contacting children on gaming and livestreaming platforms, restrict romantic or sexual AI chatbots to adults, and consider measures such as overnight curfews and breaks in infinite scrolling for users under 18.

Starmer framed the measure as a big national moment. Tech companies framed it as misdirection. YouTube and Meta warned that blanket restrictions could push children into less regulated spaces. The Molly Rose Foundation and other critics argued that age verification can be worked around and that bans do not address the central machinery of harm: algorithmic amplification and addictive design.

The enforcement question is not a footnote. It is the policy. If verification is weak, the ban becomes adult theater. If verification is strong, it creates a new infrastructure for proving age across the internet, with obvious privacy stakes. If children migrate to darker platforms, the state may have reduced official exposure while increasing actual danger.

Still, the move has force because it names a public exhaustion that polite online safety rules have not satisfied. Parents are tired of being told to manage billion-dollar behavioral systems one household at a time. Britain has now chosen to make the companies carry the burden. Whether it has chosen the right tool is a separate question, and the one that will decide whether this is law or merely catharsis.

The World Cup produced the kind of moment that makes cynicism look briefly underdressed. Lionel Messi, nearly 39, scored his first World Cup hat trick, helped Argentina beat Algeria 3-0 in Kansas City, and tied Miroslav Klose’s men’s tournament career scoring record with 16 goals. AP reported that he wiped tears after the first goal and left to a standing ovation after the third.

So yes, people are still watching. They watch because every so often the game turns time into a prop and lets an aging genius make the calendar look foolish.

But this World Cup is also becoming a study in interruption. FIFA’s new hydration breaks, introduced for summer heat across the United States, Canada and Mexico, are drawing backlash for changing the rhythm of matches. AP reported criticism from former players and current stars who say the stoppages kill momentum and resemble American timeouts. Fox cuts to commercials during the breaks. Telemundo does not. That detail tells you almost everything about the cultural fault line.

There is also the price of admission. PBS reported that FIFA priced group-stage tickets from $140, with regular seats for the July 19 final outside New York listed up to $8,680 and hospitality seats up to $73,200. After criticism, FIFA offered 130,000 tickets at $60 to national federations for regular supporters. That is a concession, but not a philosophy.

Then there is the tournament’s geopolitical residue. Iran’s team complained that it had to leave the U.S. shortly after its match against New Zealand and return to Mexico rather than recover overnight. U.S. officials said that was the agreed process, with the team allowed in the day before a match and expected to leave the evening after. A visa problem for winger Mehdi Torabi was later resolved, AP reported.

The tournament still has the match. It still has Messi. But the frame is louder now: commercials inside stoppages, ticket ladders that treat devotion as a luxury product, border rules shadowing the schedule. The World Cup has not become passe. It has become contested real estate for attention, money and state power. The ball is still the point. It is no longer allowed to be the whole story.

The British social media ban is satisfying because it points at the correct villain and then throws the nearest chair.

Children are not failing to self-regulate. Adults built machines to defeat self-regulation, funded them with advertising, optimized them by attention capture, and then acted baffled when parents could not neutralize the whole thing between dinner and bedtime. In that sense, Starmer’s move has moral clarity. It tells platforms that the burden belongs to them.

But a ban on under-16 users risks becoming the kind of policy that flatters adults while training children in evasion. The internet is not a pub with a bouncer. It is a mesh of devices, borrowed credentials, VPNs, group chats, gaming spaces and look-alike services. If the enforcement is loose, the law becomes symbolism. If enforcement is tight, Britain builds an age-verification regime with privacy hazards that will outlive the panic that produced it.

The better target is the design layer. Ban infinite scroll for minors. Ban algorithmic recommender systems that amplify self-harm, sexualized content and extremism to children. Restrict direct messaging from unknown adults by default. Make platforms liable for predictable design harms. Require independent audits, not annual trust-and-safety poems.

Age gates may have a role, especially for sexualized AI chatbots and adult content. But childhood will not be restored by pretending teenagers can be made absent from the social internet. The point should be to make the machinery less predatory, not merely to push young users into unofficial rooms.

The state has finally stopped whispering politely at the platforms. Good. Now it needs to aim at the business model, not the birthday.

The most useful question about AI is no longer whether it is big. It is big. The useful question is who gets the economy it makes.

Jensen Huang’s Texas pitch is the strongest version of the optimistic case: AI will not just automate screens, it will build factories, upgrade supply chains and create physical work. Good. Let that case stand where it can be counted. How many permanent jobs? What wages? What public subsidies? How much power and water? Which firms capture the margin? Which towns carry the infrastructure burden?

The old oil economy had rigs, tankers, refineries, unions, ports, spills, taxes and wars. People could see enough of the apparatus to fight about it. The AI economy prefers to appear as magic: a model answers, a chip stock rises, a mayor gets a ribbon cutting. But the material chain is still there: fabs, lasers, substations, rare inputs, cooling systems, data centers, grid upgrades and public incentives.

Readers are right to suspect that oil prices alone no longer explain the economy’s center of gravity. AI is becoming the asset class, the industrial policy and the labor question at once. That makes it too important to leave to keynote language.

Every major AI infrastructure project should publish a bill of materials for civic life: expected jobs by wage band, energy demand, water demand, tax abatements, local housing impact, training commitments, ownership structure and projected productivity gains. If the public is being asked to host the machine, it deserves more than promises that the future will be exciting.

Opportunity is not a plan. It is a word executives use before the accounting starts. Start the accounting.