The government wanted the exchange sold. It wanted the auction logic in the ad server opened up. It wanted a contingent divestiture of what was left of the server if the first two did not work. Brinkema rejected all three.

What she ordered instead is "most of the parties' proposed behavioral remedies" — the phrase is from the short preview filing, because THE ACTUAL ORDER IS SEALED FOR FOURTEEN DAYS while everyone requests their redactions. So the world knows the shape of the punishment and not one specific term of it, and will not know until roughly 16 September.

Her reasoning, from the remedy phase, is the part worth reading twice. Ripping the exchange out might hurt the small publishers who use the ad server for free. A buyer for the exchange — pick your giant — could introduce complications of its own. And behavioural fixes would be QUICKER than a breakup dragged through years of appeal.

Read that last one slowly. THE LENGTH OF THE APPEALS PROCESS WAS AN ARGUMENT FOR THE SMALLER REMEDY. The bigger you are, the longer you can litigate, and the longer you can litigate the more a court has to weigh the delay against the cure.

One of the government's own witnesses in the remedy trial put the hole in it plainly: what is a publisher supposed to do if it wants a different ad server and still wants this company's buy-side demand? He does not think the ruling answers that. A competing exchange operator took the other side, saying behavioural remedies should be a more expeditious path to the same level field.

Both parties now have thirty days to submit a jointly proposed final judgment. That is the document that will actually bind anybody.

The question at the top of the petition is dry enough to hide what it does: whether the 2010 Dodd-Frank Act preempted states from regulating sports bets placed inside their own borders, if those bets are offered on a market registered with the federal derivatives regulator.

Here is why it is live NOW rather than eventually. In April the Third Circuit held, 2-1, that the Commodity Exchange Act preempts New Jersey's gambling law — a clean win for the exchange that sued the state. The dissenting judge said the plain truth out loud: the products on the platform are sports bets. Then last week the Ninth Circuit, which covers Nevada, held that the same federal statute LIKELY DOES NOT preempt a state's sports-contract rules.

So the same federal law now means two different things depending on which side of the country you place the trade from. A spokesman for the federal derivatives regulator said the split "call[ed] out for resolution by the Supreme Court." The state attorney general's line is blunter: these firms "have no right to offer their sports bets without following state law."

A cert petition is not a grant. There are dozens of cases running the same argument, and the court may well wait for a few of them to finish before it wades in.

The copy is friendly. Already on the in-house wallet? No action needed, your next payout lands there. Not on it? Then enrol, and the word optional does not appear anywhere.

International creators stay on the old processor. This is a United States cutover.

And it is five days ahead of a deadline that has been on this board's ledger for a week: on 7 SEPTEMBER the legacy revenue-sharing programme is retired outright, and everyone earning from it is moved to a rewards programme weighted toward original content.

The sweetener is real, and it is not nothing: payouts become instant, instead of biweekly runs with a thirty-dollar floor. If you are a mid-sized account, the difference between money in a week and money now is a working-capital change, not a perk.

But notice the ORDER OF OPERATIONS. The rails moved first, then the programme dies. By the time the thing you were paid by is switched off, the thing that pays you is already in-house.

On 27 AUGUST an executive order directed the federal geographic naming service to rename Lake Ontario as Lake America. The two giants complied: one of them explained, reasonably enough, that it updates names to follow official government sources. The other did not comment and appears to follow the same rule.

The old mapping company posted on a social platform that it would not change the name. That was the entire strategy.

WATCH THE RANK MOVE. 128 on 28 August. 92 the next day. Number 3 on Monday 31 August, and number one in navigation. Number one overall by Tuesday.

One measurement firm estimates 632,000 United States downloads since the order. The company itself claims over 1.5 million installs across the United States and Canada. More than half of its year-to-date American downloads — 56 percent — happened in the six days from 27 August. Day-over-day downloads up 123 percent; against the same week last year, up more than fiftyfold.

And the general manager's quote is the one that should frighten the incumbents, because it is not about downloads: people are filing feature requests and bug reports, in the app reviews and on social, and a small team is shipping against them around the clock.

Somebody just acquired a couple of million engaged users for the price of a post declining to do something.

If you only know the outline, here is what the outline misses. Wilson took a voice that could have lived comfortably inside the jazz canon and walked it into folk, into blues, into country, into pop, into whatever the song needed, and she did it without ever sounding like a crossover project. She sang Robert Johnson. She sang Hank Williams. She sang Joni Mitchell and the Monkees and made them sound like they had been waiting for her.

TIME MAGAZINE CALLED HER AMERICA'S BEST SINGER IN 2001, which is exactly the kind of pronouncement that ages badly and in this case did not.

What made it work was the low end. That voice sat further down than a jazz vocal is supposed to sit, unhurried to the point of insolence, and it dragged whatever material it touched down into the same register — the song stopped being a standard and became a room she was standing in.

There is a generation of vocalists working now who sound like they were raised on the space she opened, and most of them have never had to explain why a jazz singer would record a delta blues, because she already answered it.

"I'm used to doing the impossible," he said. "Part of the problem is I'm a contrarian. I'm fascinated by doing things that people say can't be done. This one almost broke my will."

Three hundred thousand square feet. More than thirty galleries. A building shaped like something that landed rather than something that was built, sitting next to a 1923 colosseum in Exposition Park, designed by Ma Yansong of MAD, with landscape by Mia Lehrer. It opens to the public on 22 SEPTEMBER.

WHAT IS ACTUALLY IN IT IS THE PROVOCATION. Norman Rockwell in quantity. Frida Kahlo and Diego Rivera. Maxfield Parrish. Frank Frazetta. Comic book art. Western paintings. Photography. And a gallery of full-size props from the films that paid for the whole thing.

The founder's stated thesis: honour the illustrators. "Most of the stories are myths; they're not real, but the illustrations make them real. A picture is very powerful, and the stories are very powerful, and you combine them together, and those two elements create a society."

That is a real argument, and it is one the American museum system has spent a century refusing — the hierarchy that puts fine art above illustration is precisely the hierarchy that kept Rockwell out of the rooms Kahlo hangs in. Here they are on the same floor plan.

Nineteen days. That is the interval between the death of the most famous living artist in the world and the first dated posthumous exhibition on a New York calendar, and this desk means that as a measurement rather than an accusation.

THE NUMBERS UNDERNEATH IT are why the interval is that short. At her death she was the most expensive contemporary female artist alive, and in 2023 the best-selling artist of any kind at auction, at 80.9 million dollars. That same year she licensed her work to a French luggage house that put her designs on a 150,000 dollar bag and a 50,000 dollar surfboard, installed an animatronic version of her in a New York store and a colossal bust of her head on a Paris flagship.

And the artist herself had not been seen in public since 2017, had stopped receiving visitors at the Tokyo psychiatric hospital where she lived for nearly four decades by 2019, and last visited New York in 2013. The studio kept producing throughout, run by a small team of assistants.

So the final decade being sold as her most extraordinary is a claim about a body of work made by a woman in her late eighties and nineties, largely unseen making it, mediated entirely by a studio and a market that had already turned her into a global consumer brand.

The film profiles an advice columnist during her defamation litigation against the sitting president. Mainstream distributors and streamers would not buy it. The director has publicly blamed what she calls the Trump Effect and says buyers ghosted her.

So the film went around them. Over a hundred theatres. A sold-out run at one New York cinema that extended to five weeks. Screenings in more than ninety cities across thirty states. The demand was never the problem.

AND HERE IS THE STRUCTURE, because the structure is the story: FIVE COMPANIES did the work one distributor normally does. The financier-producer. An indie studio that took non-theatrical, eventised theatrical and organic social. A firm that ran educational access for schools and colleges. A physical theatrical outfit. And a screening platform where communities host their own showings, which is where the film has now launched to rent.

The head of that platform and the head of the indie studio both told the trade press the obvious: no major streamer was coming, so they pooled and shipped it themselves. "The distribution landscape, it's fundamentally broken and ripe for some change," one of them said.

The economics are deliberate. The platform's revenue share is small compared to the big transactional stores, so more money reaches the filmmakers. And the coalition keeps the controls: they set the length of the premium window — probably through year end — they can add the big platforms later, and the film stays in cinemas the whole time.

Run the timeline, because the timeline is the whole signal. NOVEMBER 2025: files a confidential draft registration statement with the securities regulator, days after raising 800 million dollars at a 20 billion dollar valuation, including 200 million from a market-making giant. MARCH 2026: plans frozen, blamed on market conditions. SEPTEMBER 2026: not before the second quarter of 2027.

That is at minimum a twenty-two month wait so far, and the new floor puts it near three years from filing to a listing that may still not happen.

AND THE BUSINESS IS NOT THE PROBLEM. Second-quarter adjusted revenue of 508 million dollars, up 17 percent year on year. Funded accounts up 42 percent to 6.6 million. Forty billion dollars of assets on the platform. It has spent the freeze buying things — a derivatives venue closed in May, a stablecoin payments platform in July, a wallet infrastructure business agreed after that — turning an exchange into a financial services company while nobody could buy the shares.

It is not alone in the waiting room. An asset manager, an Ethereum software developer and a hardware wallet maker have all postponed. The 2026 the industry expected — a wave of listings following two successful debuts last year — did not arrive.

THE MATHS IS THE STORY. Founded 2024. Raised 29 million dollars total — a 6.2 million seed and a 23 million Series A. Valued at 157 million before the sale, per a private-markets database. Sold for 500 million. Roughly a three-times step-up on the last private mark, in under twenty-four months, for a company automating password resets and app access requests.

AND THE DETAIL THAT SHOULD MAKE EVERY BOARD MEMBER SIT UP: among the angel investors in the acquired company was the ACQUIRING COMPANY'S OWN CHIEF EXECUTIVE, personally. He also gave the statement announcing the deal, describing the acquisition as giving his platform "the arms and legs to deliver autonomous security outcomes across the entire enterprise." The acquirer declined to comment on the price.

This is the buyer's SEVENTH acquisition of 2026, after an observability platform at 3.35 billion and another security startup at 400 million.

The founder had done this before — his previous company, a code-security platform, was bought by a payroll and HR company. The customer list here included a payments company, a surveillance-technology firm and an AI lab. The nearest competitor hit a one-billion-dollar valuation last December on a 75 million dollar Series B.

Exhibitors said first-day sales were slow, some noting a slight improvement on last year. Korean dealers said pieces had been PUT ON HOLD and would take another day or two to close, because buyers are cautious. Overall price points ran from five to six figures, with occasional four-figure sums.

What actually sold on day one, at the top: a new Adrian Ghenie oil at 1.1 million euros — 1.28 million dollars — from an Austrian dealer, who also placed a 2024 Antony Gormley cast-iron at 750,000 pounds. A Korean gallery sold multiple works by the late painter of water drops, one at 1.1 million dollars.

AND THE MOOD IS THE FINDING, NOT THE TOTALS. "Gone are the partygoers, spectators, and speculators," is how the trade described it. The dealers who stayed are talking in a completely different tense — one of them, five years into building a Seoul space run by a local team, said the more established you are, the more you are within a grid, and the more you build up an audience. He also said the Korean audience is among the most sophisticated in Asia and that there is nowhere else with so many private museums. "People here demand culture."

An adviser working between Taipei and New York put the actual open question best: younger audiences are showing up at exhibitions and fairs in a way that was not visible a few years ago, and it has NOT converted to buying. "But that's exactly the layer I'm watching."

Calling it a launch is not generous, it is accurate: it is the first robotaxi service open to the public in London, ahead of the American giant everyone assumed would get there first and ahead of the Chinese one now partnered with the other ride-hailing app. And London is only the SECOND CITY IN EUROPE with autonomous taxis, after Zagreb — where a Croatian operator runs Chinese technology with a supervisor on board.

The ride-hailing company's head of product told a wire agency the plan is phased and small: vehicles "in the order of dozens to start". Fares are the same as a normal ride.

AND DURING A DEMONSTRATION FOR THAT WIRE ON TUESDAY, THE OPERATOR TOOK BACK CONTROL at a junction near the Palace of Westminster. The startup's head of business development said the quiet part with admirable directness: "This is exactly why we start with safety operators behind the wheel. It's to ensure that people are safe when they're driving."

Now the part that actually dates the technology. Neither company would give a date for removing the driver. The American rival says it intends to carry passengers with nobody behind the wheel BY THE END OF THIS YEAR, subject to government approval. A third operator is waiting on guidance from the city's transport authority for a 2027 service.

And the transport authority says the guidance is coming "shortly" — and that NO OPERATOR HAS YET BEGUN THE APPROVAL PROCESS for fully autonomous vehicles.