Plutonic Rainbows

The last date anyone attached to this was this week. On 6 August a leaker named a release candidate and claimed Astra was days away. It's the 15th, nothing shipped, and next week rests on the same grade of sourcing.

OpenAI named Astra its next major model on 1 August, in a post about mathematics. Whether it ships as GPT-6, GPT-5.7 or something else is undecided, which is why the markets disagree with themselves. Polymarket gives Astra 14% by month's end, while the GPT-6 contract for 21 August has fallen to 2%, from 14% eleven days ago. One prices a release, the other a name.

Anthropic's half isn't a separate coin flip. The Chinese outlet 36kr reported in late July that Fable 5.1 has finished internal testing, that staff already use it, and that Anthropic will ship within hours of GPT-6. One report, unconfirmed. If it's roughly right, both releases share a single trigger.

Anthropic could do it on no notice: Fable 5 on 9 June, Sonnet 5 on 30 June, Opus 5 on 24 July, each live the day it was named. That's capability rather than intent, and neither lab wants to launch inside the other's news cycle. A conditional release makes that 2% a ceiling, not a starting point, with the hold story still to be true underneath.

Alain Wertheimer wanted something new to wear. That is the origin Chanel's own chairman gave for the fragrance his house quietly put into its boutiques in 1987, and it accounts for nearly everything strange about how Bois Noir was sold. No department stores. No gift-with-purchase, a mechanic Wertheimer loathed. No campaign worth the name. Speaking to Women's Wear Daily that November, he was unusually direct about why houses so rarely bothered with anything difficult: "People want success too quickly. From a fragrance point of view, people like what their memory tells them is good. Memory tells you it's good if it relates to something else you like. When a fragrance is completely different, they're lost."

Jacques Polge had been working toward a masculine scent organised around a single material, sandalwood, since the middle of the decade. The reference point was in the house already: Bois des Îles, Ernest Beaux's sandalwood for women from 1926. Bois Noir was conceived as its masculine counterpart, and the listed ingredients make the ambition plain enough. Mandarin, lavender, rose, coriander, Bourbon vanilla, sandalwood, rosewood, ambrette seed. Rose and vanilla, sold to men, in a market whose commercial leaders at the turn of the decade were Polo, Aramis, Drakkar Noir and Obsession for Men. Nothing about that formula was calculated to move volume, and the distribution model confirms that nobody at Chanel expected it to.

The packaging carried the same refusal. A stopper cut from real wood rather than moulded to imitate it, on a black box with a fine white edge, at a moment when men's fragrance was reaching for lacquer and chrome.

There is almost nothing to weigh as reception, and this is the part the collector literature skips. No contemporaneous criticism, no sales figures anyone has published, no period reviews to set against each other. A perfume sold through a handful of shops, unadvertised, over a run the sources cannot agree the length of, does not generate reception in any useful sense. What gets reported as the reception of Bois Noir is almost entirely retrospective, assembled from the 2000s onward by people writing about decants poured from bottles bought twenty years earlier. That is not worthless, but it is testimony about a heavily aged liquid, gathered from a self-selecting group who had already decided the thing was worth hunting.

Nobody can fix the dates either, and the disagreement is more telling than any single date would be. Now Smell This says the launch was discreet and the fragrance was taken out of production quickly. Raiders of the Lost Scent puts the circulation at 1987 to 1989. Perfume Intelligence records it as discontinued in 1990, on the rebrand, and files it under the flat verdict "not a commercial success." A Fragrantica feature dates the Paris boutique sale to March 1986 and so disagrees with everyone about the beginning as well. Four accounts of a Chanel product from within living memory, and they cannot fix either end of its life. That is what withdrawal without announcement looks like in the record: a scent can be sold by one of the most documented houses in fashion and still leave a gap where its dates should be.

What followed is well documented, because Chanel chose to document it. In April 1990 the fragrance returned as Égoïste, the name bought from Nicole Wisniak's photography magazine because the marketing department found Bois Noir insufficiently provocative for a global launch. Paris was papered with posters carrying the single word JE before anyone knew what was being advertised. Jean-Paul Goude directed the commercial, which used no male models at all and took the Golden Lion at Cannes. It was the largest campaign the house had ever mounted for a men's fragrance, and it went into department stores, which is to say it inverted every decision made about Bois Noir three years earlier.

Then Égoïste's sales dropped rapidly, and Chanel's explanation to the trade press was that it had proved too sophisticated for the American market. That line deserves the same scepticism as anything the collectors say. It flatters the house, blames the buyer, and arrives with no figure attached in any account I can find. Whatever the cause, the house acted on a different diagnosis: by 1993 the name was on Égoïste Platinum, a clean citrus-herbal developed with consumer test groups, which Chanel had not used before. Platinum became by some distance the better known of the two, and still is. Wertheimer had said in 1987 that people like what their memory tells them is good, and six years later his own house was convening test groups to find out what that was.

The myth now runs in the opposite direction to the evidence. Bois Noir is treated as a lost masterpiece, prices have gone somewhere absurd, and fakes circulate freely, usually slightly evaporated Égoïste decanted into a vial. Yet the most careful comparison anyone has published came back the other way. Kevin at Now Smell This gathered samples from two separate original bottles, plus a third from a trusted vendor as a control against a bad decant, and reported that Bois Noir is not that different from Égoïste: richer and denser, and recognisably the same fragrance that is still in production today. I quoted that review here years ago.

So the collector economy has spent three decades grading a working version. A draft can beat the finished piece, and I've no quarrel with the people who prefer this one. But the premium is being paid for rarity, and rarity here was manufactured by a distribution decision rather than earned by the liquid. Compare it with Kouros, launched six years earlier and written off in a newspaper headline ten days after its launch party, which then sold continuously for forty-five years while the oakmoss and civet it was built on were regulated out from under it. Collectors chasing a 1981 bottle are chasing materials that genuinely are not in the current formula. Bois Noir was tweaked and renamed rather than gutted by regulation. Richer and denser is a real difference, and I can't tell you whether it comes from the formula or from thirty-nine years in glass, and neither can the market, because almost every comparison being made is a 1987 bottle against a current one, where age and formula move together. That ambiguity is the whole trade. What age does to a fragrance is doing at least some of the work connoisseurship claims for itself, and a 1987 sandalwood that has sat in glass this long will read deeper and stranger than a current production bottle whatever was in it to begin with.

"Acid nostalgia" is Lawrence English's term for the version of memory that got repurposed over the past decade, the one that flattens the past into a single clean image and uses it to cancel whatever might have come next. He sets it out in the liner notes to The Rest Is My Ghost, released on his own Room40 label on 7 August. This nostalgia, he writes, "seeks only singular recollection", and has become "a methodology at best, and a weapon at worst, used to erode the past and project forward a collapsed and unimaginably sanitised version" of what came before. The record closes a loose trilogy begun with Wilderness of Mirrors in 2014.

Which means it arrived pre-argued. Uncut traces the concept back to Mark Fisher and Adam Curtis, The Skinny calls the album a sonic essay on memory, and Simon Kirk's Sun 13 review spends its opening on the societal weight of nostalgia before it reaches a note. I came to it the same way, through the writing around it, so this describes a reception as much as a release. A thesis travels faster than a drone, and English has handed everyone a very good thesis.

The track titles are better evidence than the liner notes. "Sodium Vapour Halo", "Metabolic City", "Aerial Cable", "One Line Sky": infrastructure, and the sliver of sky between towers. Sodium is the sharpest of the four, because that light is being switched off. Councils have been swapping the orange for white LED since 2016, the best part of half a billion pounds of it, so the haze is becoming a colour people remember rather than see, warm, uniform, flattering to nothing. Exactly the sanitised single image English says the weapon is built out of. The sleeve makes the same argument in a picture, that collapsing grey canyon, two cities tipping toward each other across a pale jagged seam, printed like a bad photocopy.

On paper the cast is comically stacked: Chris Abrahams and Tony Buck from The Necks, Norman Westberg from Swans, Thor Harris, Chuck Johnson's pedal steel, Mats Gustafsson, The Australian Voices, Madeleine Cocolas. Reviewers hunting for them come back with very little. Backseat Mafia's John Parry gets Westberg's guitar twisting in the swell of "Acid Nostalgia", and not a great deal else. English buries his collaborators the way a city buries its older streets.

The disagreement is at the end. Kirk hears English climbing out of the void onto new land. The Skinny hears "not despair, but certainly not resolution", and Backseat Mafia hears the closing bell of "GhostType" as either decline or restoration and declines to choose. One reading of recovery against two of irresolution is not much of a split until you notice what the album is about. A record arguing against singular recollection has produced criticism that cannot settle on a single memory of it.

Design or luck, and I would guess design, because the same refusal runs through the sequencing. No climaxes, no release, pieces rolling into one another so the thing has to be taken whole. The weapon English names works by making the past small and certain, and he has built fifty minutes that will do neither.

From London this evening the Moon covers 91.3% of the Sun, with the partial phase running 18:17 to 20:06 and maximum at 19:13 BST. That sounds like most of an eclipse. It isn't. The last sliver of photosphere keeps the sky blue and the shadows sharp, and without a forecast telling you to look, you might miss it entirely. Cardiff and Belfast reach 93%, Dublin 94%, and not one of them sees the corona.

Madrid and Barcelona make the point harder. Both fall a fraction outside the path and settle for 99%, and the corona stays inside the path. Totality lands on Greenland, western Iceland and northern Spain, plus a corner of Portugal, and the longest anyone gets is 2 minutes 18 seconds, 45 km off the Icelandic coast. Spain gets it low and gold just before sunset, which is either the best version of the thing or a good way to let one hill ruin your evening.

Fifteen million people will stand inside the path; another 283 million get 90% or better. The second group isn't having a smaller version of the first group's evening, which is the whole reason anyone books a flight to Bilbao for two minutes of dark. Europe's last mainland totality was 1999, and Britain's next is 2090.

Chloé Bakalar left OpenAI in July, about eleven months after joining from Meta, where she had been chief ethicist since 2019. The Financial Times reported it this week, and by that account she was the company's only dedicated ethicist, with no replacement named.

OpenAI's response is the part worth keeping. Asked to confirm there is now no ethicist on staff, a spokesperson said AI ethics "doesn't live with one owner or team" and that ethical considerations are "deeply embedded into the model building process driven by a number of teams across research." As a principle that's fine, and probably right; a lab where ethics is one person's job has already lost. As a description of who actually does what, it can't be audited from outside. The principle also arrived only once the owner had gone, and the job itself was created in August 2025.

The exodus framing built up around this deserves less confidence than it's getting. The FT piece is single-sourced, neither Bakalar nor the company has given a reason, and her LinkedIn still lists the role, so even the shape of the exit is thinner than the coverage around it. Reading a departure as a warning shot, when nobody involved has said anything, produces a story that writes itself and then gets cited as though someone reported it.

What needs no motive is the shape of the org. Superalignment stopped existing as a standalone team in 2024, Mission Alignment was disbanded earlier this year, and its former head Josh Achiam left in July, the same week as Safety Systems head Johannes Heidecke. Each of those sat outside the teams shipping the models. Chief Research Officer Mark Chen told Wired the goal now is safety work integrated with frontier-model development, earlier and closer to launch decisions.

That might genuinely be the better design, and the distinction worth holding onto is between responsibility and authority. Spreading responsibility across every team is good practice. Authority to stop a launch is a different thing, and it has to sit somewhere a deadline can't reach it. The behaviour that needs catching is rarely the behaviour anyone asked for, which is why whoever raises it needs a job that survives the raising. A reviewer inside the build shares the shipping deadline; a dedicated one doesn't, and that's most of the point of having one.

Anthropic will start marking what Claude produces. Text gets an imperceptible watermark woven into the words themselves, and supported files (.svg, .png, .jpg) get signed C2PA provenance metadata. Models launched in the EU on or after 2 August 2026 carry it from launch, older models are being retrofitted during the grace period the law allows, and the marking is applied at the model level worldwide, across the API, Claude Code, Cowork and the cloud partners. None of it is optional and none of it is visible.

The driver is Article 50 of the EU AI Act, in force since 2 August, which requires machine-readable marking of synthetic content that is "effective, interoperable, robust and reliable as far as technically feasible". Anthropic signed the accompanying code of practice, and the code is voluntary while Article 50 is not. Marking obligations are arriving through other doors too. California bundled watermarking guidance into its procurement rules back in March.

The C2PA half is close to decorative. The manifest comes off with a re-save, a format conversion, a screenshot, or an upload to a platform that rewrites metadata on the way in, and The Verge notes it is sometimes stripped by accident. Marks written into the pixels themselves hold up far better. OpenAI's checker was solid enough that people used it two days ago to identify an anonymous model on Arena. Metadata riding alongside a file has never managed that.

The text watermark is the consequential half, and it is the one with nothing published behind it. Anthropic doesn't name the scheme, hasn't released technical documentation, and hasn't shipped the detection tool it says it is working on. The properties it does claim are that the mark rides inside the text, so copy-paste carries it, while heavy editing, paraphrase or translation can leave nothing detectable, as can a passage that is simply short.

Those two properties invert who gets caught. C2PA fails indiscriminately, coming off for anyone who re-saves the file. The text mark is selective, and selective in the wrong direction: someone pushing Claude's output through a second model to pass it off as their own clears it without trying, while someone who pasted Claude's tidy-up of a paragraph they wrote themselves keeps it. The careful user stays marked.

Anthropic is straight about the limits. A detection means content may have been processed by Claude, not that Claude wrote it, and no detection proves a human did. That is the sentence that matters, and it sits in a help-centre article. Detection tools are promised rather than delivered, so nothing has gone wrong yet. Once they arrive, the result will be read out in misconduct hearings and HR meetings by people who never opened the caveat, treated as a test result rather than a hint, the same way plagiarism-checker percentages got treated. Marking the output is the right call. Shipping the marks a long way ahead of any means of interpreting them is not.

An anonymous image model called mona-lisa-1 turned up on Arena today, and several people independently flagged it as OpenAI's within hours. @AiBattle_ posted the first sighting this morning, having run the outputs through OpenAI's own verification tool and got a hit. That matters more than it sounds: OpenAI watermarks its images with SynthID via a partnership with Google, and the checker looks specifically for a watermark originating from OpenAI. One commenter noted that the other accounts passing the story around had left that detail out.

So the provenance is reasonably solid. What the model actually is stays wide open, and the early reads are strikingly muted. @synthwavedd ran comparisons through the day and called it a slight improvement in realism and noise over GPT-Image 2, not a big jump. @WolfRiccardo posted raw outputs two hours earlier and framed it as rumour. On Reddit the verdict was that "it doesn't seem deeply different".

Blind arena testing won't settle it. Everyone is running different prompts with no seed control and no fixed comparison set. @synthwavedd reads the "-1" suffix as a promise of further checkpoints, which fits April, when three tape-themed codenames ran on Arena before shipping as GPT-Image 2 on the 21st.

No outlet has covered mona-lisa-1. For now it's a few thousand views on X and a Reddit thread that can't tell whether it's looking at anything new.

Two more frames have surfaced from the Alberta Ferretti autumn/winter 1992 show in Milan, alongside the five I wrote about in July. The question I keep being asked about this collection is whether it laid any groundwork, and the honest answer is narrower than either yes or no: nothing in these clothes anticipates the house Ferretti became. Whether it mattered at the time is a separate question, and one I'll come back to.

What the name means now is settled. Lisa Armstrong, writing in the Times, pinned the signature as a kind of Valentino look for a younger, slightly more bohemian woman, and the house itself sells that reading back in the language of chiffon, slip dresses and lace. Transparency and drift. It is the register that put Uma Thurman in crimson at the Oscars in 2000 and never really left the red carpet afterwards.

None of it is in these pictures. Helena Barquilla comes down the runway in a red wool jacket cut short and nipped, the shoulder built up, chinchilla at the collar and swallowing both cuffs so her hands emerge from grey fur in black leather gloves. Under it run brocade trousers in the same red, gold urns and cartouches printed the length of the leg, and the shine on them is doing all the work the jacket refuses to. That's the third chinoiserie piece I've now seen from this one show, after the mandarin-collared brocade jacket and the beaded suede column, so the theme was running properly through the collection rather than turning up once as a novelty. Heavy, opaque, structured, warm. The opposite of drift.

Whether Milan noticed, I can't tell you, and the reason is itself worth recording. Vogue Runway's archive does reach back to 1992, and its entire fall season that year comes to three houses; Ferretti is not among them. The Fashion Model Directory indexes the show and then answers 403 on every page that would name the cast, and Bloomsbury's catalogue entry sits behind a subscription wall. What survives in the open from these years is the business rather than the clothes: the Milan showroom at Palazzo Donizetti in 1994, New York two years after that, Bergdorf Goodman and Neiman Marcus by mid-decade. An expansion is easier to record than a collection.

So the clothes are free to be of their moment. The purple suit is period Milan almost to the point of parody: violet bouclé jacket and short skirt, a magenta leather shell zipped to the throat underneath, black suede gloves again, and a belt of gilt baroque medallions wide enough to be structural. Gold hardware over bright wool was general across the seasons either side, Krizia included. Set against the chiffon that came later, the violet and the gilt read less like an early draft than a different house altogether.

Two greens, and they are not meant to match. The storm coat is coated to a wet shine, cut wide enough through the sleeve to swallow whatever sits underneath, and fastened with gold snaps instead of buttons so the front stays flat. The double-breasted jacket beneath it lands in the same pistachio register but matte, tailored close, four gold domed buttons running down it. Wearing the two together is the proposition: one surface throws the studio lights back and the other absorbs them.

Under the greens Gail Elliott has powder blue on powder blue, a silk shirt with a white contrast collar and flapped patch pockets buttoned to the throat, over trousers in the same blue, dry and heavy. No print anywhere on the body. The pattern is quarantined into two scarves, one worn as a headband, one knotted at the neck, both in the same blue-green-orange floral. Six months later the winter book would put her in tartan and carriage-print silk from the collar down.

Then the gold: a plaque buckle with two hearts flanking a capital E, baroque earrings set with blue stones. Escada spent the eighties making gold the argument rather than the trim, and on this page that is where the volume has gone. The fabric behaves. The hardware does not.

In April 1993 the chief designer Michael Stolzenburg told the Los Angeles Times that "Escada will do no grunge", and sold retailers an easier, more youthful autumn, on the grounds that younger customers no longer felt it was the right time to be showy. Some of that easing had already happened in the cloth before he described it, and the autumn book then arrived in tartan and gold medallions anyway. Nobody was going to mistake this coat for a flannel shirt. But the restraint the eighties never managed is in the plain blue trousers, not in anything the house said about them.