The most consequential economic announcement of Tuesday morning did not come from a central bank or a finance ministry. It came from the Pacific Ocean.

El Niño is strengthening, and the U.S. Climate Prediction Center now puts the chance of a very strong event during the Northern Hemisphere fall and winter above 90 percent. An experimental NOAA forecast expects a historically strong peak later this year. Across Southeast Asia, the warning is already physical: stressed reservoirs, farmland fires, weaker fish catches and the prospect of drought spreading through some of the world’s most important rice, palm-oil and manufacturing regions.

Weather forecasts often arrive as colored maps whose uncertainty makes them easy to admire and ignore. Food systems do not have that luxury. A failed harvest is converted into lower export supply, higher feed costs and government intervention. Those costs travel. Rice becomes dearer in an importing country that has never seen the dry field. A poor corn crop raises the price of meat months later. Reduced hydropower output creates more demand for fuel precisely when energy markets are already carrying a war premium.

This El Niño therefore matters less as an isolated climate episode than as a force multiplier. Asian governments have spent reserves cushioning households and businesses from higher energy and fertilizer costs associated with the Iran war and disrupted shipping. Fiscal protection is finite. The stronger the ocean warming becomes, the more governments must choose among subsidizing food, defending currencies, buying fuel and financing disaster response.

The sequence is worth watching carefully. First comes water stress: cities restrict use and farmers receive less irrigation. Then crop estimates decline and traders begin bidding for substitute supplies. Exporting governments, anxious about domestic prices, can impose restrictions that make the global shortage worse. Importers build stockpiles. Consumers encounter the result much later, one package and restaurant bill at a time.

There is a political temptation to treat every price increase as a discrete villainy—greedy supermarkets, careless governments, panicked traders. Some of those accusations may be justified. But the chain begins with a genuine reduction in abundance, transmitted through energy, shipping and finance. Punishing the last seller does not refill a reservoir.

There is also a temptation to overstate certainty. El Niño rearranges risk; it does not dictate identical weather everywhere. Some regions receive flooding while others dry out, and local outcomes depend on timing, infrastructure and policy. The responsible claim is not that every harvest will fail. It is that governments and markets have been given unusually strong notice of a broad, correlated shock.

That notice creates a test. Grain reserves can be audited before panic buying. Import contracts can be diversified before harvest numbers collapse. Water restrictions can be imposed before taps run low. Cash assistance can be designed before food inflation forces families to reduce meals. The boring work of preparation is cheaper than emergency procurement and almost always less visible.

Hurricane Polo, now a major storm off Mexico’s Pacific coast, supplies the day’s spectacular image. El Niño is the larger story because it is not one storm. It is a season-long alteration of the background conditions in which storms, droughts and markets operate.

For readers, the useful indicators will not be the adjectives attached to the ocean. Watch reservoir levels in Southeast Asia, crop revisions for rice and palm oil, fertilizer prices, export restrictions and the currencies of large food importers. That is where a Pacific temperature anomaly becomes household arithmetic.

The weather report is no longer adjacent to the business pages. It is the first draft of them.

World leaders arrive at the United Nations on Tuesday with an unusually crowded list of emergencies and unusually little agreement about their causes.

The wars in Ukraine and Iran remain active. Gaza, Sudan, Congo and Myanmar demand attention without commanding it. Disruption around the Strait of Hormuz has become a security and cost-of-living question far beyond the Gulf. Climate shocks and artificial-intelligence governance are competing for diplomatic bandwidth with conflicts measured in casualties rather than model capabilities.

President Donald Trump is scheduled to address the General Assembly and hold meetings with British Prime Minister Andy Burnham, Japanese Prime Minister Sanae Takaichi and Venezuela’s acting president Delcy Rodríguez, according to the White House schedule reported by The Associated Press. The program is consequential, but the bilateral meetings around it may matter more than the set-piece speech. The U.N.’s annual gathering remains one of the few places where governments that cannot agree on a communiqué can still share an elevator bank.

The institution’s problem is not a shortage of topics. It is that the largest powers increasingly describe the same crisis in incompatible language. Maritime attacks can be framed as terrorism, deterrence, retaliation or resistance depending on the speaker. AI can be an existential hazard, an industrial opportunity or a sovereignty project. Climate policy looks different to a fossil-fuel exporter, a drought-stricken importer and a country trying to finance data centers.

That makes procedural achievements worth separating from substantive ones. A new working group is not a ceasefire. A leaders’ statement is not an enforcement mechanism. A meeting between adversaries is evidence of contact, not reconciliation.

The week should be judged by narrower tests: whether maritime talks produce operational rules, whether humanitarian access expands in any active war, whether governments attach dates and responsible agencies to AI proposals, and whether climate-finance promises identify actual money. The speeches will furnish language. The useful news will be found in the verbs that survive after the delegations leave.

A federal appeals panel ruled Monday in favor of Missouri’s new congressional districts, allowing a map backed by President Donald Trump to advance while leaving the ultimate decision to the U.S. Supreme Court.

The case is the latest stage of a long-running dispute over the boundaries used to elect Missouri’s U.S. House delegation. Supporters of the map argue that the state acted within its authority. Challengers contest the districts’ legality and have sought judicial intervention.

The appeals decision is important but not final. Until the Supreme Court acts, candidates, parties and election administrators must plan around a legal instrument that could still change. That uncertainty has practical consequences: campaign organizations decide where to spend, prospective candidates assess whether to run, and local officials prepare ballots and voter information.

The court’s eventual ruling will determine more than the fate of particular boundary lines. It will help establish how lower courts should evaluate state redistricting decisions and when federal judges may intervene. For now, the precise fact is narrower: Missouri won this round, and the challenged map remains alive.

Readers should watch the Supreme Court docket rather than campaign declarations for the next meaningful development. The operative questions are whether the justices take emergency action, what timetable they impose and which map election officials are instructed to administer.

The final major obstacle to Paramount’s acquisition of Warner Bros. Discovery is no longer a prohibition. It is a production schedule.

Twelve states and Hollywood writers agreed Monday to settle lawsuits challenging the deal, which The Associated Press values at $81 billion. The settlement still requires a judge’s approval, but it clears a central path toward combining Paramount, Warner Bros., CBS, CNN, HBO Max and Paramount+ under one corporate roof.

The unusual part is the remedy. Paramount must release 30 films annually during the first two years, including 20 wide releases, and 32 annually during the next three years, including 21 wide releases. At least four films a year must be independent productions. Missing the quota could trigger a $30 million payment for each absent film and require the company to divest Miramax.

The company also committed at least $1.5 billion in additional U.S. production spending over five years compared with its 2025 level. A further $47.5 million will support training and career development for workers displaced by the merger. Other provisions address cable negotiations and oversight of the editorial independence of news operations.

These are more concrete commitments than the soft assurances that often accompany mergers. They also reveal the limits of the settlement. A required number of movies is not the same as vigorous competition between independent studios. Two executives choosing 30 releases can satisfy a quota while making a narrower range of bets than two rival companies would have made separately.

The enforcement language nevertheless matters. Film production is volatile, release dates move easily and corporate promises decay rapidly after a transaction closes. A per-film penalty and a possible asset divestiture give the commitment a balance-sheet consequence.

Consumers should watch subscription prices, theatrical windows and the volume of canceled projects after the merger, not merely the annual film count. Workers should watch where the additional production spending occurs and whether it offsets consolidation elsewhere. Journalists at CBS and CNN should watch who appoints the promised oversight bodies and what authority they actually exercise.

The settlement has attached measurable obligations to a transformative deal. It has not demonstrated that the transformation is harmless.

Alibaba has answered the eve of a U.S.-China summit with three large numbers: a chip it says is three times faster than its predecessor, an artificial-intelligence model planned at five trillion to 10 trillion parameters, and more than 20 gigawatts of global cloud data-center capacity targeted by 2032.

Chief executive Eddie Wu unveiled the Zhenwu V900 accelerator at the company’s Apsara Conference in Hangzhou, calling it China’s most powerful AI chip. The company said it is mobilizing resources across models, semiconductors and data centers to satisfy demand for AI services.

The strategy is vertically coherent. Alibaba does not merely want to train a frontier model; it wants to design the chips, operate the cloud and sell access to the resulting systems. That integration matters in China, where access to leading foreign accelerators is constrained and domestic companies have strong incentives to control more of their compute supply.

But parameter count is a plan, not a performance result. Larger models can require more memory, electricity, training data and engineering time without producing proportionate gains for customers. The commercially meaningful tests will be cost per inference, reliability, latency, energy efficiency and the ability to make useful software rather than demonstrations.

The 20-gigawatt target may ultimately prove more revealing than the model’s size. It describes a physical estate of power connections, cooling systems and capital expenditures. Alibaba is proposing to compete not only in algorithms but in the industrial system required to run them.

The timing makes the announcement geopolitical theater as well as product strategy. President Trump and Chinese leader Xi Jinping are expected to meet this week with AI competition, trade and export restrictions among the sources of tension. Alibaba’s message is that China intends to pursue frontier scale with a domestic stack.

Investors should withhold applause for the nouns. The chip needs independently observed deployments. The model needs evaluations that measure capability and cost. The data centers need permits, power and paying workloads. Ten trillion is an impressive number. It is not yet an impressive product.

A finished movie that nobody is allowed to watch acquires a peculiar glamour. It becomes purer than a released film because nobody can complain about the third act.

“Coyote vs. Acme,” Dave Green’s long-shelved collision of animation and live action, is finally reaching cinemas this month. Its arrival is part movie release, part jailbreak. The film became famous before audiences could judge it, after Warner Bros. declined to release it and the decision turned a corporate accounting maneuver into a public argument about whether completed art should be treated as disposable inventory.

That history creates a nearly impossible critical burden. Viewers are not merely being asked whether the jokes work. They are being invited to vindicate everyone who objected to the shelving. A merely good family comedy must now carry the symbolic weight of labor, preservation and artistic dignity.

The healthiest response is to let the film become ordinary. See it. Dislike parts of it. Argue over the performances. Permit it the same vulnerable existence as every other movie. The point of saving an artwork cannot be to place it beyond criticism; it is to make criticism possible.

Its release also lands as Hollywood prepares for a new corporate order. Paramount’s proposed acquisition of Warner Bros. is moving forward under a settlement that mandates dozens of annual releases. That agreement counts movies. “Coyote vs. Acme” is a reminder that the essential distinction is not between 29 and 30 titles. It is between a work that can enter public memory and one that remains an entry in a spreadsheet.

The film may be delightful, disappointing or both. It has already supplied a durable image of the modern studio system: Wile E. Coyote finally catches something, only to discover that the box containing it belongs to the accounting department.

Alibaba’s proposed model may contain as many as 10 trillion parameters. This is a fact of the announcement and almost no description of the product.

Technology has done this before. Camera companies spent years teaching buyers that more megapixels meant better photographs, even when lens quality, sensor size, processing and the person holding the camera mattered more. The number succeeded because it transformed a complex judgment into a larger-is-better label that could sit on a box.

Parameter count performs the same magic for artificial intelligence. It converts an argument about reasoning, reliability, efficiency, data quality and usefulness into a number with enough zeroes to look like destiny. The company gains a headline before it has supplied a model.

This does not mean scale is irrelevant. Large systems have produced real gains. It means that scale has become most rhetorically powerful precisely where it is least informative. A model twice as large may not be twice as capable. It may be dearer, slower and harder to operate. A smaller system may be superior for a particular job.

The antidote is not another universal benchmark. It is a bill of particulars: What can the system do that existing models cannot? What does a completed task cost? How often is it wrong? How much energy does it consume? Can outsiders reproduce the claimed result? Which customers will pay for it without subsidy?

Ten trillion parameters may turn out to be a genuine engineering achievement. Until those answers arrive, it is better understood as the world’s largest sticker on a box nobody has opened.

The settlement clearing Paramount’s acquisition of Warner Bros. has the satisfying texture of enforcement. There are annual film quotas, wide-release quotas, independent-film quotas, spending commitments and a $30 million penalty for every missing movie. Someone has finally put teeth into a merger promise.

The teeth are attached to the wrong animal.

Antitrust law is supposed to protect competition. The settlement instead permits two major studios to combine, then attempts to reproduce the output that rivalry might have generated by giving the enlarged company a production syllabus. This is regulation as fan fiction: the competitive market disappears, but officials write a version in which the merged company behaves as though it still had an adversary.

Thirty films can be inexpensive obligations, contractual contrivances or sincere works. A release count says little about budgets, marketing, creative risk or bargaining power. The merged studio will naturally optimize for whatever the agreement measures, because every institution does. If the rule counts titles, expect titles.

The possible Miramax divestiture and per-film fines make this stronger than a ceremonial pledge. The additional $1.5 billion in domestic production spending and worker fund are material. Yet even good enforcement cannot turn a quota into competition between independent decision-makers.

The settlement’s own complexity is the confession. Officials are trying to specify outputs, theatrical reach, independent sourcing, labor support, cable negotiations and editorial safeguards because the merger concentrates power across studios, streaming services and newsrooms at once.

If the deal closes, the public should insist on every promised film and every promised dollar. It should also remember that receiving the compensation package is not the same as avoiding the injury. A movie quota is a useful covenant. It is a poor substitute for another studio capable of saying no, yes or something entirely different.