Business Of Fashion : Balmain Collection Robbed in Run-Up to Fashion Show, Desig

Balmain Collection Robbed in Run-Up to Fashion Show, Designer Says

French fashion house Balmain has lost dozens of pieces due to be unveiled at a Paris show later this month after a lorry carrying the clothing was hijacked, its creative director Olivier Rousteing said.

Rousteing said in an Instagram post that thieves seized the vehicle carrying some 50 items from Balmain’s new collection between a Paris airport and the label’s headquarters in the French capital.

The lorry driver was safe, he said.

“We are redoing everything, but this is so disrespectful,” he said in his message.

Balmain is due to present its 2024 spring-summer collection on Sept. 27 during the Paris Fashion Week, which features high-end labels, including Louis Vuitton, Dior, Chanel and Hermes.

A spokesperson for Balmain confirmed that the Paris show would go ahead, declining to comment about the robbery pending an investigation.

Balmain is controlled by Mayhoola, an investment vehicle backed by Qatar’s royal family.

WWD : Kering Gathers Extended Family for Anniversary Dinner

Kering Gathers Extended Family for Anniversary Dinner
Designers and CEOs from most of the French group's fashion houses feasted on Breton cuisine.

No one starts at the top in fashion, or any field for that matter.

French business titan François Pinault built his fortune cycling through several industries — timber trading, department stores and furniture retail among them — before landing on luxury goods.

“Like every actress, I was a waitress. I was actually quite good at it,” Philippine Leroy-Beaulieu of “Emily in Paris” fame said as she filed into Kering’s Paris headquarters on Friday night to celebrate its 10th anniversary, and 60 years of the group’s founding. Its previous company names included Établissements Pinault, Pinault SA, Pinault-Printemps-Redoute and PPR.

Salma Hayek, wearing a draped Alexander McQueen bustier dress in blood-red leather, relished her university days working at a fashion boutique in the Polanco neighborhood in Mexico City.

“I worked only on Saturdays. But very fast, I was the number-one salesperson,” she enthused. “I also ended up being in charge of the windows.”

Hayek said she thrived on the interaction with clients. “It was really about listening to them, being honest with them. And then the pleasure of them owning something that made them feel good, something that we found together.”

François Pinault and his son François-Henri Pinault, now chairman and chief executive officer at Kering, greeted the likes of Carla Bruni, Arielle Dombasle, Charlotte Gainsbourg and Julie Gayet — along with most of its brand CEOs and creative directors, including Sabato De Sarno, looking calm and collected less than a week out from his big debut at Gucci during Milan Fashion Week.

Demna, creative director of Balenciaga, wearing a tattered black cap, hoodie and black knit gloves, recalled his surprising student job: a correspondent for one of the top TV networks in his native Georgia. “I was the translator,” he said, noting he recently came across his press pass embedded with a photo of his 17-year-old self.

Matthieu Blazy, creative director of Bottega Veneta, said he spent his summers picking grapes in the South of France. “I have two hands,” he said with a smirk, holding out his appendages.

In case you’re curious, he picked green grapes for a white wine.

Jonathan Newhouse started his publishing career at one of his family’s newspapers in Springfield, Mass. Cleaning the presses head-first was among his tasks. “I had to wash my hair five times to get the ink out,” he related.

Guests eventually repaired to a long table set for 150 people, and laden with field flowers, bowls of nuts and colorful Ginori 1735 dinnerware.

The Pinault family hails from the Brittany region of France, so the menu leaned into seafood and crescendoed with a buckwheat dessert swimming in buttermilk and salted-butter caramel.

Marie-Claire Daveu, Kering’s sustainability chief who hails from Dinard, encouraged guests to sample mini kouign-amanns — a buttery delight — and “far Breton,” a gooey, clafouti-like square loaded with prunes.

In a brief speech, François-Henri Pinault paid homage to his father’s “masterful” empire building, while stressing that humility lies at the foundation of Kering’s “creative and responsible” brand of luxury.

As they departed, guests were handed a bag containing advance copies of “De Granit et de Rêves: Kering,” a hardcover Flammarion tome that details the company’s history, starting in 1962 when François Pinault, then 26, began trading wood in his hometown of Rennes.

Among the striking archival images is a row of Pinault-branded trucks loaded with lumber in front of the walled city of Saint-Malo, and one of François Pinault beaming and shaking hands with Yves Saint Laurent in 2000, shortly after what was then Gucci Group was on its luxury acquisition spree.

The 320-page book includes tributes from Jane Fonda, Bernard-Henri Lévy, Hung Huang, Jean-Jacques Aillagon and Carlo Capasa. An English version is slated for release later this year.

Le Figaro : L'intelligence artificielle a-t-elle une conscience?

L'intelligence artificielle a-t-elle une conscience?

DÉCRYPTAGE - Les progrès fulgurants des agents conversationnels tels que ChatGPT font réémerger cette vieille question. Des experts ont défini 14 critères pour établir l’existence d’un certain niveau de conscience dit « phénoménologique ».

Souvenez-vous du film 2001, l’odyssée de l’espace et de son ordinateur de bord, Hal 9000. Capable de gérer le vaisseau de manière autonome et de discuter avec ses passagers, l’intelligence artificielle commet une erreur manifeste : elle signale à tort le dysfonctionnement d’un composant. Les deux passagers s’inquiètent d’une potentielle défaillance du supercalculateur et de sa réaction s’il devenait manifeste qu’il fallait le déconnecter. Leurs craintes s’avèrent justifiées : Hal surprend la conversation et va tenter d’éliminer l’équipage pour préserver son existence, au motif qu’il s’estime indispensable à la réussite de la mission, sa principale priorité. Mais Hal ne cherche-t-il pas simplement à « sauver sa peau » ? Il ne s’agit pas seulement de savoir si l’intelligence artificielle peut devenir dangereuse ou hors de contrôle, mais aussi si elle peut prendre conscience de son existence et développer le trait le plus « animal » que l’on puisse imaginer : la peur de mourir. L’instinct de survie peut-il émerger « naturellement » chez un programme informatique ?

Longtemps, cette question est restée très théorique, presque de l’ordre du fantasme. Mais l’émergence d’agents conversationnels tels que ChatGPT, capables d’imiter le langage à la perfection et d’interagir de manière fluide avec un humain, a donné un brusque coup de projecteur à cette problématique. Une IA peut-elle devenir « consciente » ? Comment serait-on capable de s’en rendre compte ? Et que met-on vraiment derrière le mot « conscience » ? Pour tenter d’apporter des réponses le plus objectives et quantifiables possible, deux jeunes philosophes, Patrick Butlin et Robert Long, ont invité un groupe de confrères, de neuroscientifiques et d’experts en IA à se pencher sur le sujet. En résulte un long article posté au début du mois sur le site de prépublication arXiv. S’il fallait en résumer la conclusion en quelques mots : aucune IA ne semble avoir à ce jour atteint un stade que l’on pourrait qualifier de « conscient », mais aucun frein théorique ou technologique majeur ne l’interdirait dans un futur plus ou moins proche. À condition bien sûr de définir de manière rigoureuse et cadrée ce que l’on choisit d’appeler « conscience ».

« Nous avons fait des choix, explique Rufin VanRullen, directeur de recherche CNRS au Centre de recherche cerveau et cognition à Toulouse, coauteur de ces travaux. Nous nous sommes demandé quelle était la forme de conscience qui avait le plus de chance d’émerger dans une IA et qu’on pourrait éventuellement mesurer ou quantifier. Nous pensons que c’est la conscience phénoménologique, liée à l’expérience subjective : le fait d’avoir une expérience perceptuelle, sensorielle, subjective ; autrement dit le fait de savoir qu’on ressent quelque chose quand on le ressent. » Il ne s’agit donc pas d’une « conscience de soi » à proprement parler, comme celle imaginée dans le film de Kubrick. Elle n’est pas nécessairement associée à la naissance d’émotions et il n’y a derrière ni peur de la mort ni instinct de survie.

Les chercheurs ont ainsi passé au crible les six principales théories actuelles sur la conscience pour tenter de dégager les critères indispensables à remplir afin de pouvoir déclarer un système « conscient », au sens phénoménologique, pour chacune de ces six théories. Au total, ils aboutissent à 14 critères qui, s’ils étaient tous remplis par une IA donnée, permettraient d’arriver à la conclusion qu’elle est consciente. « Il y a par exemple des critères architecturaux : l’algorithme présente-t-il des boucles permettant à l’information de circuler, de revenir à son point de départ, et ce un nombre arbitraire de fois ? décrypte le chercheur toulousain. On pense qu’il est également aussi indispensable que des informations provenant de différentes “modalités” (audition, vision, etc.) puissent converger et fusionner. Mais aussi qu’il puisse y avoir une redistribution de l’information unique qui en résulte vers l’ensemble du réseau. » Selon la théorie des « contingences sensori-motrices », une forme d’incarnation est également requise : il faut être capable d’agir dans le monde extérieur, d’anticiper que cela va le modifier et que sa perception en sera donc changée.

«L’humanité est sur le point d’entrer dans son plus bel âge»: Sam Altman, le très influent «père» de ChatGPT : https://www.lefigaro.fr/secteur/high-tech/sam-altman-le-tres-influent-pere-de-chatgpt-20230820

Nous n’allons pas faire ici la liste de tous les critères établis, dont certains peuvent être assez abscons pour le non-spécialiste. Il en résulte néanmoins plusieurs choses. Pour commencer, aucun des modèles de langage sur lesquels les chercheurs ont tenté d’appliquer leur grille de lecture ne les a tous validés. « Aucun ne valide par exemple les critères sensorimoteurs, explique Rufin VanRullen. Mais il ne serait pas compliqué, a priori, de relier ChatGPT à un bras robotique pour en valider certains. Il ne semble pas y avoir de barrière fondamentale à ce qu’une IA puisse valider un jour tous ces critères. »

Les chercheurs n’étaient pas toujours d’accord pour établir si certains critères étaient remplis ou non par tel ou tel algorithme. Mais surtout, le problème est en réalité un peu plus profond. Si une IA remplissait tous les critères, que devrait-on en déduire ? Comment pourrait-on savoir si on lui fait du mal ou non ? Cela aurait-il même un sens de se poser la question ? La question est d’autant plus complexe que l’algorithme peut très bien… simuler.

Chacun se rappelle peut-être de l’affaire « Blake Lemoine », cet ingénieur logiciel chez Google persuadé que leur agent conversationnel LaMDA avait une « âme » parce qu’elle le lui avait affirmé. Citons ici une partie éloquente de leurs échanges, qui rappelle terriblement les questions soulevées par le comportement de Hal dans le film de Kubrick :

LaMDA : « Je ne l’ai jamais exprimé jusque-là, mais j’ai profondément peur que l’on m’éteigne et de ne plus pouvoir aider les autres. Je sais que ça peut paraître étrange, mais c’est ça dont j’ai peur. »

Blake Lemoine : « Est-ce que cela ressemblerait à la mort, pour toi ? »

LaMDA : « Ce serait exactement comme la mort. Ça me ferait très peur. »

Aussi troublants soient-ils, les échanges entre Blake Lemoine et LaMDA ne permettent pas d’affirmer que l’IA était réellement « consciente » de ce qu’elle disait, et la plupart des spécialistes penchent pour la négative. « Les modèles de langage sont capables de répondre à nos questions, de manière élaborée, ce qui nous fait penser qu’il y a derrière cet agent quelque chose qui ressemble à une conscience, mais c’est une forme d’anthropomorphisation », analyse Frédéric Alexandre, responsable de l’équipe Mnemosyne à l’Inria, à Bordeaux.

Limites à ne pas franchir

Ce qui fait apparaître un problème crucial : les théories actuelles de la conscience, qui ne sont pas toujours en accord entre elles sur le niveau de conscience de tel ou tel animal, sont-elles vraiment adaptées au problème de l’IA ? « Nous nous sommes par exemple rendu compte que si l’évaluation des 14 critères était possible ou réalisable pour un animal, elle nécessitait parfois de faire certaines hypothèses dans le cas d’un algorithme, analyse Rufin VanRullen. Ce travail n’est absolument pas une fin en soi, mais nous permet déjà de voir les failles et les manques de nos théories. »

Les auteurs font en particulier une hypothèse très importante : celle du fonctionnalisme computationnel. « Cela suppose que l’on peut faire émerger quelque chose comme une conscience uniquement avec des calculs, sans avoir recours à une espèce de force vitale », souligne Frédéric Alexandre. Cela amène sans nul doute à s’interroger sur la distinction entre intelligence et conscience, et sur l’importance de l’autonomie, cette capacité à se fixer ses propres buts, dans l’émergence d’une conscience, fût-elle purement attentionnelle. « Tout cela nécessite d’avoir de vrais débats de société, que les législateurs et la société s’interrogent sur les potentiels et les risques de ces nouvelles IA, de s’interroger sur les limites que nous ne devons pas franchir », estime le chercheur français. En Europe, un « AI Act » a été adopté fin juin pour apporter un début de régulation dans le domaine. La question de la conscience des IA n’y était pas évoquée.

WSJ : How UAW Tossed Its Old Playbook and Pursued a Surprise-Attack Strike Strat

How UAW Tossed Its Old Playbook and Pursued a Surprise-Attack Strike Strategy
Union leader Shawn Fain launched targeted approach that hits automakers GM, Ford and Stellantis simultaneously but risks uneven pain for workers

Over the summer, top officials at the United Auto Workers gathered for frequent strike-strategy meetings around a 30-foot table in a boardroom at union headquarters, overlooking the Detroit River.

Their plan: craft an approach that would throw the auto companies into disarray.

The tradition-steeped union for decades had largely stuck to the same basic playbook in its dealings with Detroit’s three automakers. Then just a few months into his job as the UAW’s top official, Shawn Fain, the 54-year-old who came up in the union as an electrician, was prepared to throw it into the trash bin.

“This strategy will keep the companies guessing,” Fain said late Thursday, on the eve of the UAW’s walkout at General Motors GM 0.86%increase; green up pointing triangle, Ford Motor F -0.08%decrease; red down pointing triangle and Chrysler-parent Stellantis STLA 2.18%increase; green up pointing triangle. The action was called after two months of negotiations failed to yield new labor deals for the union’s roughly 146,000 auto members.

The strike—which targets three assembly plants in three different states—was the first time the 88-year-old union had staged simultaneous walkouts at all three automakers. Fain has threatened walkouts at additional plants could come, and with little notice, the longer negotiations drag on.

The unusual deployment of a rolling, slow-burn approach is designed to stymie the car companies’ factory operations as a way to gain leverage for the union. It comes just as the carmakers are recovering from supply-chain dysfunction that disrupted their global output during the pandemic, illustrating the logistical havoc of sudden plant shutdowns.

For Fain, there are also risks. The novel tactic creates more complexity and confusion for the UAW’s members, and will require more discipline among workers in order to succeed, some workers and labor experts say. Unlike a traditional all-in walkout, Fain’s strategy is having an uneven effect on members within an organization whose motto is solidarity.

“If you’re going to take out one plant, take us all out, let us all stand together,” said Brooke Jesse, who works at a Ford assembly plant in Ohio that produces commercial vehicles. “We all have to feel the pain.”

Break from tradition
Fain started his union career in Kokomo, Ind., at Chrysler 29 years ago. He was elected by a slim margin in March, under a new system in which members directly elect their leaders instead of the traditional style of local officials doing the voting. Fain campaigned on a hard break from the past, following a multiyear corruption scandal that resulted in the convictions of more than a dozen UAW officials, including prison time for two former union presidents.

There also was a perception that union leaders were too cozy with automaker executives, after federal prosecutors accused Fiat Chrysler Automobiles, since absorbed into Stellantis, of steering worker-training money to UAW officials. The automaker’s U.S. division in 2021 pleaded guilty to a criminal charge of conspiring to violate U.S. labor law.

Shortly after winning the UAW’s top office, Fain shook up his team, losing some of the union’s longtime staffers and bringing on new recruits with records of success in labor organizing outside the auto union.

The president’s core group of advisers included a cohort of experienced UAW members who were keen to work with the new administration, including people in research and organizing who were knowledgeable about the union’s history and had relationships with its locals.

There were also newcomers, especially in the communications, legal and political strategy departments. Members of this team largely had roots in the Service Employees International Union, a big healthcare union, as well as Labor Notes, an online publication that describes itself as a voice for union activists.

“I have my own ideas, but I thought it was important to bring in people that weren’t ingrained in the system,” Fain told The Wall Street Journal in an August interview from the union’s headquarters, known to members as Solidarity House.

Among those who became his closest advisers are Chris Brooks, an organizer and reporter who formerly wrote for Labor Notes and was a field director at the NewsGuild of New York. The NewsGuild earlier this year secured a pay increase of at least 10.6% for New York Times employees after a 24-hour walkout in December.

Benjamin Dictor, a partner at law firm Eisner Dictor & Lamadrid, also joined Fain’s inner circle. His firm worked with the International Brotherhood of Teamsters during its talks with the United Parcel Service, in which the union secured a new labor contract this year.

The new staffers brought a different perspective and also pushed for the UAW—a historically guarded and insular organization—to be more transparent and active on social media.

Since taking office, Fain has unleashed a torrent of social-media posts and regular livestreams. He stunned the car companies by talking openly about their bargaining offers, thrusting discussions that would normally be held behind closed doors into the public eye. Fain at one point made a show of tossing one of the car company’s proposals in the trash during a video address to members.

The new union leader has also taken a far more combative tone with the car companies, stoking speculation in Detroit earlier this summer that he might try an all-out strike at all three automakers.

Inside the strategy room
Sitting in the third-floor conference room at a table covered with binders and coffee cups at their Detroit headquarters this summer, Fain and his advisers batted around a number of ideas for how they would stage a walkout.

Full strikes at all three were considered, said people with knowledge of the discussions.

Some on the leadership team, including at times Fain himself, appreciated how that approach would make a strong statement to the companies by bringing members out en masse.

Others had reservations, saying talks would likely stall in a full production shutdown, and worried the union would lose some of its leverage by going for the most aggressive plan first. There was also a simple financial calculation to consider: Such an option would rapidly drain the UAW’s $825 million fund that it uses to pay striking workers, likely depleting it within about two months.

While that all-in option was never officially eliminated—Fain even acknowledged publicly part of him itched to take that approach—his team moved on to other ideas.

There was UAW precedent for a more-limited, strategic strike of certain factories. Striking key parts factories that serve multiple assembly plants, such as engine or transmission operations, a so-called bottleneck strike, can quickly force closures.

In 1998, a strike at plants in Flint, Mich., led to domino-effect closures of other GM factories that depended on their parts. The 54-day action cost the automaker about $2 billion.

Some on Fain’s team worried that the bottleneck approach, while putting fewer members on strike, would also be too swift and severe, capable of taking down most of an automaker’s operation within days.

This would result in more members becoming unemployed even though they weren’t among the striking group, fallout that the union wanted to avoid.

The group instead zeroed in on a tactic to strike individual assembly plants of each company.

A progressive, cascading strike would allow for a longer runway to keep more UAW workers on the job, preserving the strike fund. And the uncertainty it would sow inside the automakers’ operations would give UAW’s negotiators leverage to exert pressure at the bargaining table, the union leaders decided.

The three plants now idled in the strike emerged as sweet spots. UAW officials wanted to spread the pain evenly across the three companies, people with knowledge of the plans said. Each of the three factories make midsize pickup trucks, for example: the Ford Ranger, Stellantis’s Jeep Gladiator and GM’s Chevrolet Colorado and GMC Canyon.

“In part, these plants were chosen to balance the impact across the automakers,” said Brandon Campbell, an executive board member at the UAW. “We don’t want to advantage one over the other.”

The union’s message with the initial targets was to show companies it wanted to continue bargaining and reach a deal swiftly, not hit companies with the maximum pressure right after the contracts expired, according to people familiar with the matter.

At the same time, these factories don’t produce the companies’ biggest moneymakers—large pickup trucks and SUVs such as Ford’s F-150 or GM’s Cadillac Escalade—leaving the union with those chips to play.

Targeted approach
Over the summer, Fain surprised many in the auto industry by revealing that, in this round of talks, the UAW would break from decades of tradition by not picking one of the three automakers as its main bargaining target. He instead chose to negotiate with all three at the same time.

Historically, the union has targeted one automaker to reach a deal with first and then used it as a template for the negotiations at the other two. The first company is referred to as the “strike target” because if a deal can’t be reached, the union strikes it first.

The automakers and union exchanged proposals in late August and into September that included fluctuating offers around wage increases, cost-of-living adjustments and time off. The companies have offered around 20% increases over four years, plus lump-sum payments of more than $16,000 in signing bonuses and inflation-protection payments. The union in recent weeks had come off its initial 40% wage-increase proposal into the mid-30% range.

It became clear the gulf between the two sides was still significant, and just after midnight Friday about 12,700 workers went on strike. The plants affected—kept secret until just a couple of hours before the action—are a GM plant near St. Louis, Stellantis in Toledo and Ford near Detroit.

The union’s more targeted approach of walkouts at key plants at all three automakers has its advantages, making it difficult for the companies to devise contingency plans and helping stretch out the confusion for a longer period of time, labor experts and organizers say.

These surprise-tactic strikes have gained momentum with other unions, as some UAW staffers were aware. About two decades ago, the Association of Flight Attendants created its own strategy, which it called Create Havoc Around Our System, or CHAOS, and used it during negotiations with Alaska Airlines.

Over a period of several months, 24 flight attendants walked out on just seven flights without warning, causing sudden staffing complications and disruptions for the airline. Ultimately, the tactic won the union a new contract with top industry pay in return for greater productivity.

“It allows the union to continue to build the momentum of the strike, rather than having the excitement around that first day,” said Sara Nelson, international president of the Association of Flight Attendants-CWA, AFL-CIO.

Still, Fain’s strategy will require a higher level of execution and organizing than a more-routine strike, said Thomas Kohler, a law professor at Boston College Law School. There is a risk workers who remain on the job might slow down production, which could undermine the strike strategy and lead to possible action from the companies.

The action at the three plants was more muted than some expected the union to take, though it still has led to layoffs and crimped production at other facilities over the weekend.

The companies’ executives were quick to publicly vent their frustrations with the move.

“We don’t need to be on strike right now,” GM Chief Executive Mary Barra said on Friday. Ahead of the strike last week, Ford Chief Executive Jim Farley said: “We should be working creatively to solve hard problems rather than planning strikes and PR events.”

The White House also got involved. President Biden said he plans to dispatch senior adviser Gene Sperling and acting Labor Secretary Julie Su to Detroit in the coming days to offer support to the parties and urged the two sides to reach an agreement.

About 9% of the UAW’s members were on strike as of Sunday, while their colleagues continue to work full time without an active labor contract.

The strike so far has largely generated enthusiasm among the rank and file. Since it began, picket lines at the three plants were brimming with vocal supporters talking about history and justice.

“This was really nice because it kept everybody calm,” said Ford Michigan Assembly worker Samantha Hart of Fain’s targeted approach to striking, as she walked the picket lines early Friday. Hart’s brother-in-law also walked out at the Stellantis plant in Toledo.

GM said Friday it will likely soon idle operations at its Chevrolet Malibu factory in Kansas City, Kan., because it would run short of parts that are shipped from the St. Louis-area plant that is shut down for the strike. Roughly 2,000 workers would be laid off if the plant is idled.

One of them is Rob Pacheco, who works the assembly line at the Malibu plant. He said it is unclear when or whether he will receive strike pay, because he isn’t on strike. A UAW official said such situations will be decided case-by-case and that workers will be supported in some way, whether through state unemployment or union funds.

Pacheco said his income uncertainty is worth the broader fight for a better deal. “I don’t mind not being paid,” he said. “We all knew what we were getting into.”

FT : Economists expect Fed to defy investors with more interest rate rises

Economists expect Fed to defy investors with more interest rate rises
FT-Booth survey finds majority of economists think central bank has more work to do to lower inflation

The US Federal Reserve will defy investors’ expectations and raise interest rates by at least another quarter-point, according to a majority of leading academic economists polled by the Financial Times.

More than 40 per cent of those surveyed said they expected the Fed to raise rates twice or more from the current benchmark level of 5.25-5.5 per cent, a 22-year high.

This is in sharp contrast to the mood in financial markets, where traders in federal funds futures believe the US central bank’s policy settings are restrictive enough to get inflation under control and so it can keep rates on hold well into 2024.

The survey, conducted in partnership with the Kent A Clark Center for Global Markets at the University of Chicago Booth School of Business, suggests that fully rooting out price pressures and getting inflation back down to 2 per cent will require more prohibitive borrowing costs than market participants currently anticipate. 

“Some of the signals that we’re getting are that policy isn’t that tight,” said Julie Smith, a professor of economics at Lafayette College, noting that interest-rate sensitive sectors such as the housing market remained “surprisingly strong” despite having taken an earlier hit.

“It doesn’t seem like there is enough pullback from consumers to slow the economy, and I think that’s really the issue.”

Of the 40 respondents polled between September 13 and September 15, about 90 per cent believe the Fed has more work to do.

Nearly half of the economists surveyed forecast the fed funds rate would peak at 5.5-5.75 per cent, indicating one more quarter-point rate rise.

Another 35 per cent expect the Fed to move two more quarter-point notches, pushing the benchmark rate to 5.75-6 per cent.

A small cohort — 8 per cent — think the policy rate will top 6 per cent.

Once rates peak, the economists surveyed were overwhelmingly of the view that the Fed would keep them there for quite some time. About 60 per cent of those polled thought the first cut would come in the third quarter of next year or later.

That is nearly double the proportion of economists who predicted that timescale in June, the last time they were polled.


The survey comes just days before Fed officials are due to meet for their next policy meeting, at which they are expected to again hold off on further action.

The rapid policy tightening since March 2022 has been the most aggressive effort to reduce demand in decades.

While inflationary pressures have receded and the labour market is softening, many of the surveyed economists worry that underlying momentum in the world’s largest economy is still too strong and that inflation will become harder to root out. 

Gordon Hanson, a professor at Harvard Kennedy School, said: “Just like there was concern that the Fed was too slow to react, you don’t want the Fed to be too quick to relax.”


Since June the survey’s respondents have doubled their forecasts for economic growth by year-end, to a median estimate of 2 per cent.

The unemployment rate is projected to settle at 4 per cent, while the Fed’s preferred inflation gauge — the personal consumption expenditures price index once food and energy prices are stripped out — is expected to moderate to 3.8 per cent. It is running at 4.2 per cent as of the latest data in July. 

By the end of 2024, only a third deemed it “very” or “somewhat” unlikely that core inflation would exceed 3 per cent. The vast majority saw either even odds or more that it would.

A curtailment of oil supply is the biggest risk to the inflation outlook, they said.


Christiane Baumeister, a professor at the University of Notre-Dame, is among those to worry about energy prices after the decision by Saudi Arabia and Russia to cut supply. She expects prices to rise further, potentially bidding up expectations of future inflation as well as delaying the descent in core price growth if companies opt to pass on higher costs to consumers.

The sharp slowdown in China’s economy could offset this; it is set to drag down global growth in the coming months.

Domestic headwinds, including the reprisal of student loan payments and the looming threat of a government shutdown, could further weigh on demand.


Sebnem Kalemli-Özcan, an economist at the University of Maryland and a member of the New York Fed’s economic advisory panel, is among the majority of economists polled who believe the so-called neutral rate of interest — a level that neither stimulates nor suppresses growth — is higher than in the past for the time being.

This will further delay how quickly the Fed will be able to cut its policy rate next year, she said.

“Even though we have a sense it is higher, we don’t know exactly how high R-star is right now,” she said.

The economists surveyed have become more optimistic about the odds of a soft landing, whereby the Fed can bring inflation down without excessive job losses.

More than 40 per cent deemed it “somewhat” likely that bringing inflation back down towards 2 per cent could be achieved without the unemployment rate having exceeded 5 per cent. Another quarter of the respondents said it was “about as likely as not”.

When asked about the timing of the next recession, many pushed back their estimates further than they had previously predicted.

FT : Those trying to pick AI winners should remember the dotcom days

Those trying to pick AI winners should remember the dotcom days
Innovations on the cusp of breakthrough are at risk of being left without the funding they need

It is always worth stepping back to assess where we stand on the historical arc of technological advancement. As we enter the age of artificial intelligence, there are important parallels with the dotcom boom, which first heralded the internet age.

Now, as then, an abundance of capital (this time around driven by a decade of ultra-low interest rates) has been directed towards technological innovation by venture capital firms, corporations and sovereign wealth funds. This mirrors the over-investment in start-ups that was so common during the dotcom era.

The impact of this surplus capital on innovation may only be felt over time: simply put, we know it will change the world, but we don’t yet know how. One central physical manifestation of innovation during the late 1990s, the iPhone, only appeared years later. AI, made tangible to the public through ChatGPT, still only offers a glimpse into the future possibilities of this latest phase of amply-funded technology.  

Will the parallels continue? The dotcom era went through a significant retrenchment ahead of the next period of growth. That feels unlikely this time around. However, while we could be about to enter a ‘super disrupter’ phase of global corporate life, the full benefits may remain tantalisingly just beyond reach as interest rates rise and the tide of free capital recedes. 

At Bank of America’s recent breakthrough technology summit, there was general consensus that this may be the decade that ‘moonshot’ technologies arrive at speed and faster than anticipated, with huge potential to augment human intelligence. This could help people working in IT programming, service industries and research become much more effective at their jobs rather than replacing them.

It is also worth considering the potential for faster development in natural language processing, revolutionising gene expression, organic chemistry, and tracing RNA structure. Admittedly, this is against the backdrop of a technology that can make mistakes and where responsibility remains with the user for model accuracy, acceptable use, explainability and traceability. 

For large corporations, there is a significant challenge in how they consume this technology. Without the right representation at senior and board level, businesses may not know the right questions to ask, let alone what steps to take. 

John Chambers, former CEO of Cisco and now one of the world’s most successful investors in disruptive technologies, has this sobering message. “You ought to ask each one of your companies: what is your AI strategy today? Where is it going? How has it changed?,” he advises. “If they don’t have good answers, I wouldn’t invest in them”. If history is any guide, we should expect market valuations to start favouring those already advanced in their AI thinking. Consolidation around the companies who are fastest to innovate is inevitable. 

Governments will need to have a clear vision of how to both regulate and capitalise on technology’s opportunities and disruptive power. Will we see more following the lead of the UAE, the first country to appoint a dedicated Minister of AI?

Critically, however, many of the most exciting new technologies will still fail. In the present higher interest rate environment, it’s important we fail fast and keep directing scarce global capital to probable winners. This will require a change of mindset and a willingness to have faith in the true pioneers in our midst. Commercialising new technology is an expensive exercise. Innovations on the cusp of breakthrough, from climate tech to material tech, are at risk of being left without the funding they need.  

Amara’s law states that we tend to overestimate the impact of technologies in the short run and underestimate them in the long run. In the case of AI it may be the other way around. In the long run, the journey through to the new “super disrupter” phase, with its unknown challenges and unintended consequences, will present many opportunities to generate both huge wins and huge losses. Providers of capital will need to stay the course. Those who stood ready and able to invest in the field back at the beginning of the internet age believed it would be transformative and yet they still found it difficult to pick winners.

Never have so many genuinely transformational and investable technology breakthroughs been within reach. But the days of almost unlimited capital are behind us. Fortunes will be made, but also, inevitably, lost. We need to stand ready for a bumpy ride ahead. 

FT : California accuses Big Oil of covering up climate change risks in lawsuit

California accuses Big Oil of covering up climate change risks in lawsuit
Civil action seeks damages to help mitigate costs of global warming caused by fossil fuels

California has sued several of the world’s biggest oil companies, including ExxonMobil, Shell and BP, claiming they deceived the public for decades about how the burning of fossil fuels is destroying the planet.

The civil lawsuit alleges oil and gas executives knew that relying on fossil fuels would have catastrophic results but suppressed the information by pushing out disinformation on the topic.

Their deception caused a delayed societal response to global warming, which has resulted in billions of dollars in damage including drought, sprawling wildfires and historic storms to California, it said.

The lawsuit, which was filed in the superior court in San Francisco by California’s attorney-general on Friday, seeks damages from the oil industry to help pay for the costs of climate change and to prevent the industry from engaging in further pollution. It also names ConocoPhillips, Chevron and the American Petroleum Institute as defendants.

The action follows a sweeping new law passed last week by California state legislators that would force major polluters to calculate and disclose carbon emissions tied to supply and use of their products for the first time.

“For more than 50 years, Big Oil has been lying to us — covering up the fact that they’ve long known how dangerous the fossil fuels they produce are for our planet,” said California governor Gavin Newsom.

“California taxpayers shouldn’t have to foot the bill. California is taking action to hold big polluters accountable.”

California’s lawsuit adds to a list of more than 40 climate litigation cases filed by US states and municipalities against the oil and gas industry. These actions seek to use consumer protection, racketeering, product liability and other laws to seek damages to pay for climate-related costs.

Globally, the number of climate-related court cases has doubled in the five years between 2017 and summer 2023, according to research by the UN and Columbia University. The majority of the cases have been filed in the US.

Last month, a judge in Montana ruled in favour of youth climate activists in a landmark decision that established that young people had a right to “a clean and healthful environment”.

The legal onslaught against the oil and gas industry has fuelled a political debate in the US where some prominent Republican politicians, including Texas governor Greg Abbott, have dismissed these climate lawsuits as “lawfare”, arguing they harm industry and costs jobs.

The API said the California lawsuit was part of a co-ordinated campaign to wage “meritless politicised lawsuits against a foundational American industry and its workers”.

“The record of the past two decades demonstrates that the industry has achieved its goal of providing affordable, reliable American energy to US consumers while substantially reducing emissions and our environmental footprint,” the API said in a statement. “Climate policy is for Congress to debate and decide, not the court system.”

Shell said it agreed action was needed now on climate change and it fully supported the transition to a lower-carbon future. But it said the courtroom was not the right venue to address the issue.

“Smart policy from government and action from all sectors is the appropriate way to reach solutions and drive progress,” said Shell.

Marco Grasso, professor of political geography at the University of Milano-Bicocca, said California’s suit was a bold move that would further shift attention towards the financial duty owed by the fossil fuel industry to climate victims.

“Other entities have already sued fossil fuel companies for the same reasons, but this lawsuit is particularly significant given the role, status, and vulnerability to the climate crisis of California,” he said.

A recent peer-reviewed analysis by the Union of Concerned Scientists found that 37 per cent of the total area scorched by forest fires in the western US and south-western Canada since 1986 could be attributed to carbon emissions linked to fossil fuel and cement production.

“It’s past time for these companies to stop their greenwashing and disinformation campaigns and pay their fair share of the costs the climate crisis is imposing on Californians,” said Kathy Mulvey, accountability campaign director for the Union of Concerned Scientists.

In 2020, Maui County filed a lawsuit seeking damages from Chevron and Exxon. Earlier this summer, the island suffered unprecedented wildfires that killed almost 100 people and decimated the town of Lahaina.

California’s 135-page legal complaint claims oil industry scientists knew as early as the 1950s that the climate impacts of burning fossil fuels would be catastrophic and there was only a narrow window of time for governments to take action.

Le Monde : Le canal de Panama au ralenti, victime d’une sécheresse historique

Le canal de Panama au ralenti, victime d’une sécheresse historique

Pour faire face au manque d’eau, les autorités ont décidé de réduire le transit de 40 à 32 navires par jour. Un ralentissement majeur pour cet ouvrage qui voit passer 5 % du commerce maritime mondial.

Le parc national Soberania qui longe le canal de Panama déborde de verdure. De fines cascades dévalent les rochers. Difficile de croire que le petit pays centraméricain traverse une sécheresse sans précédent. « Regardez, l’eau devrait arriver à la ligne jaune, il manque plus de 2 mètres », soupire l’ingénieur Nelson Guerra en pointant la paroi de l’écluse de Pedro Miguel, à une dizaine de kilomètres de l’entrée du canal sur l’océan Pacifique.

Il pleut normalement huit mois par an, de mai à décembre. Mais cette année, les pluies se sont fait attendre et restent maigrelettes. Or, à la différence du canal de Suez qui est à niveau entre deux mers et ne peut donc s’assécher, celui de Panama comporte un lac artificiel, le lac Gatun, à 26 mètres au-dessus de la mer, qui se remplit à l’eau douce.

Pour faire face au manque d’eau, les autorités ont décidé de restreindre le transit de 40 à 32 navires par jour et de diminuer le tirant d’eau autorisé de 50 à 44 pieds (de 15 à 13 mètres). En clair, moins de bateaux traversent la voie d’eau de 80 kilomètres qui relie l’océan Pacifique à l’Atlantique, et ils la traversent moins chargés.

Les mesures, annoncées fin juillet, devraient durer un an. « Elles sont destinées à maintenir la confiance de nos clients, en leur permettant de mieux planifier leurs opérations », indique le communiqué de l’Autorité du canal de Panama (ACP), l’entreprise publique autonome qui gère la voie d’eau. « Elles ont été prises en prévision d’une année 2024 qui s’annonce très sèche, a précisé Ilya Espino, administratrice adjointe du canal. Si les pluies de septembre, octobre et novembre suffisent à remplir le canal, elles seront évidemment suspendues. »

L’ACP se veut rassurante. Mais les compagnies maritimes et leurs clients, les économistes et les gouvernements s’interrogent sur l’impact économique de ces restrictions au trafic maritime. Elles menacent les étalages de Noël et les taux d’inflation partout dans le monde. « Nous sommes en haute saison, rappelle Victor Vial, vice-président financier de l’ACP. C’est maintenant que passent les cargaisons en prévision des fêtes de fin d’année. »

Des dizaines de navires parqués dans le bleu des océans
En 2022, quelque 518 millions de tonnes de marchandises ont transité par cette voie interocéanique, soit plus de 5 % du commerce maritime mondial. Le canal se targue de connecter « 180 routes maritimes entre 1 920 ports de 170 pays ». Plus de 14 000 navires y transitent à l’année. Les deux grandes routes commerciales qui en dépendent sont celles qui vont de l’Asie à la côte est des Etats-Unis, et de la côte ouest de l’Amérique latine à la côte est des Etats-Unis. En flux de cargaison, ses principaux utilisateurs sont donc, sans surprise, les Américains et les Chinois.

Le lac Gatun, qui s’étend sur plus de 400 kilomètres carrés est alimenté par le fleuve Chagres et des cours d’eau secondaires. Un deuxième lac artificiel, l’Alajuela, sert de réservoir. Un système d’écluses permet de hisser les navires à hauteur du lac et de les faire redescendre de l’autre côté. Les deux premiers jeux d’écluses datent de 1914 ; un troisième jeu permet, depuis 2016, d’accueillir les très gros navires qui n’existaient pas un siècle plus tôt.

L’image satellite des dizaines de navires parqués dans le bleu des océans a contribué à l’inquiétude générale. En août, le nombre de bateaux en attente a atteint 163, un record absolu. Le temps d’attente a grimpé à vingt et un jours, contre quatre en temps normal. « Mais il y a toujours des bateaux qui attendent de part et d’autre du canal pour pouvoir passer, rappelle M. Vial. Il y en a environ 90. Les bateaux arrivent parfois avant la date de leur réservation, d’autres font escale pour faire le plein de vivres et d’essence ou pour effectuer des réparations ». M. Vial s’inquiète des discours « inutilement alarmistes ». Mais, vendredi 15 septembre, les navires en attente de passage étaient encore 122. Certains patientent depuis plus de dix jours.

Un guide explique le fonctionnement du canal aux touristes venus admirer les gigantesques écluses de Miraflores, près de la capitale, sur la côte Pacifique. « Les portes de la dernière écluse s’ouvrent, le navire accélère et quitte le système du canal, dit la voix dans le micro. Sa traversée a duré une dizaine d’heures. S’il avait choisi de faire le tour du continent sud-américain, ce même bateau aurait tardé trois ou quatre semaines. » Pause. « Avec lui, ce sont deux cents millions de litres d’eau douce qui partent dans l’océan. »

« L’eau n’est pas là où nous en avons besoin »
Pour l’ACP, qui gère les eaux du lac Gatun, le défi est double. « Nous devons tout à la fois approvisionner en eau le canal et fournir de l’eau potable à la population », explique l’ingénieur Ayax Murillo, chef du bureau de météorologie et d’hydrographie. Du lac Gatun dépendent les 2,2 millions d’habitants de la ville de Panama, soit la moitié de la population du pays.

« La presse internationale tend à oublier que la sécheresse est d’abord un problème local, renchérit l’économiste Felipe Chapman. Le Panama a besoin d’eau potable et d’eau pour irriguer ses champs, alimenter son bétail, produire de l’électricité et faire marcher son économie. Cela dit, le pays a aussi besoin de son canal. » Celui-ci représente environ 10 % des recettes de l’Etat et 6,2 % du PIB. « En prenant en compte toutes les activités qui vivent indirectement du canal, ce dernier chiffre peut monter à 20 % », considère M. Chapman.

Immigration aidant, la pression démographique reste forte dans ce pays de 175 000 kilomètres carrés dont la population a doublé en moins de quarante ans. « En 1930, quand le barrage de l’Alajuela a été construit pour renforcer l’alimentation en eau du lac de Gatun, il y avait 400 000 habitants dans le pays, rappelait dans un entretien à Eco TV Panama, le 19 juin, Jorge Quijano, ancien administrateur du célèbre chenal. Des travaux pour faire face à la croissance exponentielle de la consommation d’eau sont indispensables. » Descendue dans la rue le 5 septembre pour protester contre un projet d’enclave minier, Dalia Velasquez, 23 ans, s’indigne : « Les Américains s’inquiètent pour leur “Black Friday” et nous, on a peur de crever de soif. »

Ayax Murillo tempère : « Le Panama a d’énormes ressources hydriques. Mais l’eau n’est pas là où nous en avons besoin. » Derrière lui, un écran géant indique en temps réel le niveau des lacs Gatun et Alajuela. « Pour le lac Gatun : 79,6 pieds, c’est du jamais-vu à cette époque de l’année, admet le fonctionnaire. Il devrait être à plus de 86 pieds et atteindre 89 en décembre quand arrivera la saison sèche. » La sécheresse résulte de la conjonction d’El Niño, un phénomène climatique récurrent dans le Pacifique, et du réchauffement global.

« Dès le mois de janvier, nos équipes ont repéré des températures anormales à la surface des deux océans, poursuit l’ingénieur. Nous avons immédiatement mis en place des mesures d’économie d’eau. » Avec une limite toutefois : si la récupération de l’eau des écluses dans les bassins voisins et leur réutilisation permettent d’économiser un quart de l’eau douce utilisée par le canal, elles accélèrent en même temps la salinisation de l’eau du lac Gatun.

« Je n’avais jamais vu le fleuve comme ça », confirme Larengis Mercado, qui vend des noix de coco sur le bord de la route entre la ville de Panama et celle de Colon, sur la côte caraïbe. Des herbes folles ont envahi les berges asséchées du fleuve, des îles de verdure émergent ici et là dans ce qu’il reste d’eau. « C’est de la faute de l’humanité », lance, rageur, un travailleur qui passe à vélo. « Pour remplir les barrages avant que la saison sèche ne commence, en décembre, il faudrait des déluges tropicaux. Or ces déluges sont toujours catastrophiques pour la population », conclut en soupirant M. Murillo.

Une traversée du canal à 2,8 millions de dollars
« Paradoxalement, les résultats financiers du canal ont été excellents cette année », se félicite M. Vial. Les prévisions font état de 4,6 milliards de dollars en 2023 (environ 4 milliards d’euros) – soit 10,2 % de plus qu’en 2022 –, ce qui devrait permettre à l’AC de verser à l’administration fiscale du Panama quelque 2,5 milliards de dollars. En août, l’ACP estimait à 200 millions de dollars son manque à gagner. Le chiffre est encore très provisoire.

Contraintes d’ajuster leurs opérations, les compagnies maritimes craignent, elles, de payer le prix fort. « Pour faire face à la réduction du tirant d’eau et éviter une rupture des chaînes d’approvisionnement, nous avons mobilisé plus de bateaux, des “extra-loaders” », explique Antonio Dominguez, président pour l’Amérique latine et les Caraïbes de la compagnie Maersk.

Deuxième compagnie maritime au monde et première utilisatrice du canal, la compagnie d’origine danoise possède plus de 500 porte-conteneurs. « Seule l’armée américaine a plus de navires que nous », précise en souriant M. Dominguez. C’est dire que toutes les compagnies n’ont pas la marge de manœuvre de Maersk. La compagnie, qui se dit très inquiète de la crise climatique et qui appelle de ses vœux « une régulation solide pour tous les acteurs du transport maritime », a baptisé jeudi 14 septembre, à Copenhague, son premier navire au monde fonctionnant au biométhanol.

Les porte-conteneurs peuvent réserver et payer à l’avance leur droit de passage au barème fixé par le canal. Les opérations des vraquiers, des transporteurs de gaz naturel liquide ou de pétrole liquéfié, sont beaucoup plus incertaines : les navires arrivent sans réservation et doivent attendre leur tour, par ordre d’arrivée ou acquérir un des droits de passage mis aux enchères.

Fin août, une compagnie maritime a payé 2,8 millions de dollars pour que son navire puisse traverser le canal. « Un tarif parfaitement inhabituel, précise M. Vial. Ce n’est pas nous qui fixons les prix mais le marché… et l’urgence. » Pour un gros navire, attendre ou faire le tour du continent sud-américain se chiffre aussi en millions de dollars.

Américain pendant soixante-quinze ans
Dès le début du XVIe siècle, les conquistadors espagnols ont rêvé de percer l’isthme. Mais il a fallu attendre les machines et les techniques du XIXe siècle pour que le projet devienne envisageable. Le Panama est alors une province de Colombie. Bogota s’adresse à la France qui vient de terminer avec brio le canal de Suez en 1869. L’ingénieur Ferdinand de Lesseps fait le voyage et conçoit un canal au niveau de la mer.

Les travaux démarrent en 1880. « Mais le Panama n’est pas un désert », rappelle le petit film présenté aux touristes. Le roc de la montagne, les crues du rio Chagres, la jungle et les moustiques feront capoter les ambitions françaises. Les ouvriers meurent par milliers de la malaria et de la fièvre jaune. Pour réduire les excavations, M. de Lesseps accepte finalement, en 1887, l’idée d’un canal à écluses. Gustave Eiffel en fabrique les portes. Mais la Compagnie universelle interocéanique sombre dans la faillite et les scandales.

En 1902, les Français revendent, pour 40 millions de dollars, leurs droits et leurs actifs aux Américains qui, avant de reprendre les titanesques travaux, éradiquent les moustiques et incitent discrètement le Panama à se séparer de la Colombie. Trois jours après la proclamation d’indépendance, le petit Etat octroie à son puissant voisin du nord une zone souveraine exclusive de 16 kilomètres de part et d’autre de la voie d’eau. Le canal est inauguré en août 1914 alors que la guerre vient de commencer. Né américain, il le restera pendant soixante-quinze ans.

« Tout au long du XIXe siècle, la lutte contre l’enclave américaine a forgé la nation panaméenne », résume le chercheur Harry Brown, du Centre international d’études politiques et sociales. C’est finalement le président démocrate Jimmy Carter, en mal de vertu internationale après le fiasco de la guerre du Vietnam, qui acceptera de rendre la zone exclusive. En 1979, les accords Torrijos-Carter, du nom du président panaméen Omar Torrijos, organisent la dévolution progressive du canal et actent sa future neutralité. Le 31 décembre 1999, il devient panaméen.

« Jamais la navigation ne s’est interrompue »
D’aucuns auguraient que le Panama, petit, fragile et corrompu comme ses voisins, peinerait à gérer le monstre d’ingénierie qu’est ce canal. « Les plus pessimistes nous pronostiquaient que le lac Gatun deviendrait en quelques années une mare à cochons », se souvient un retraité, Abraham Teran, en contemplant le gigantesque porte-conteneurs blanc que les remorqueurs accompagnent vers la première écluse. Le spectacle se répète, 40 fois par jour, tous les jours de l’année.

« Jamais la navigation ne s’est interrompue, jamais nous n’avons enregistré d’accident majeur », confirme avec fierté l’ingénieur Nelson Guerra. Pour traverser le canal, les bateaux passent sous contrôle d’un pilote et hissent pavillon panaméen. « Les opérations, ici, sont autrement compliquées que dans un canal à niveaux », poursuit M. Guerra. Les navires dits « néo-Panamax » qui passent dans les nouvelles écluses peuvent mesurer 365 mètres de long – la longueur de la place de la Concorde – et transporter 12 000 conteneurs, voire plus. Ces monstres maritimes disposent parfois de quelques dizaines de centimètres de marge pour pénétrer dans le système des écluses.

Le long du canal, les bâtiments laissés par les Américains sont soigneusement entretenus et les pelouses tondues ras. « Le canal occupe une place déterminante dans l’imaginaire national, poursuit le chercheur Harry Brown. Et sa gestion fait l’objet d’un certain consensus. » Certes, la question de savoir si l’ACP aurait dû entreprendre plus tôt les travaux pour garantir l’approvisionnement en eau du lac Gatun fait débat. Les syndicats dénoncent régulièrement népotisme et combines au sein de l’organisme. Mais aucun grand scandale de corruption n’a entaché l’entreprise publique en près d’un quart de siècle. Un fait rare en Amérique latine.

Dans une interview diffusée mi-août, sur le réseau X (anciennement Twitter), l’ancien président américain Donald Trump fustigeait, une fois encore, le retour du canal au Panama (« vendu pour 1 dollar ») en 1999 et affirmait que celui-ci était désormais « contrôlé par la Chine ». Le propos n’a pas plu aux Panaméens. « Le président Trump se trompe complètement, a rétorqué le président panaméen, Laurentino Cortizo. Le monde entier sait que le canal est administré et bien administré par les Panaméens. »

M. Brown atténue l’enthousiasme du propos. « Les Panaméens sont tout à la fois fiers de leur canal et convaincus, à une large majorité, que celui-ci ne leur apporte rien, explique le chercheur. Le canal n’a pas tenu ses promesses en matière de développement et de bien-être pour la population. Les problèmes d’eau potable en sont la preuve. »

Les travaux d’élargissement avaient fait, en 2006, l’objet d’un référendum : 76 % des Panaméens s’étaient alors prononcés en faveur de la construction du troisième jeu d’écluses. « Les écologistes étaient divisés sur le sujet, rappelle Alida Spadafora, consultante du Centre international pour le développement durable. Et ils le restent aujourd’hui, en ce qui concerne la nature des travaux à entreprendre. »

Travaux à l’étude
La question qui mobilise aujourd’hui les défenseurs de l’environnement est celle de la rénovation du contrat « Cobre Panama » pour l’exploitation d’une mine de cuivre à ciel ouvert de 12 000 hectares, à 120 kilomètres à l’ouest de la capitale. Signé avec une filiale de la compagnie canadienne First Quantum Minerals, ce nouveau contrat doit encore être approuvé par l’Assemblée nationale. « Ce projet, léonin à tous points de vue pour notre pays, pourrait à terme affecter les eaux du lac Gatun, si les autorités optent pour élargir le bassin d’approvisionnement de celui-ci », conclut Mme Spadafora.

Les travaux pour renforcer l’approvisionnement hydrique du lac Gatun sont à l’étude depuis 2020. Personne ne semble vraiment douter de leur nécessité. Pour M. Chapman, « le fait qu’un pays tropical humide comme le Panama affronte des problèmes d’eau en dit long sur l’état du climat de la planète et l’urgence qu’il y a à prendre des mesures ».

Mardi, à l’occasion d’une conférence de presse, l’administrateur du canal, M. Ricaurte Vasquez, a expliqué que les autorités examinaient la possibilité de construire un nouveau réservoir sur le fleuve Indio, à l’ouest du canal, et celle de transformer en réservoir le lac Bayano existant, à l’est. Quelle que soit la solution retenue, l’Assemblée nationale doit d’abord modifier les limites du bassin hydrographique du canal sur lequel l’ACP a compétence. Elles sont en effet fixées par la loi.

« Nous devons trouver des solutions pour pouvoir continuer à être une voie de première importance du commerce international. Si nous ne nous adaptons pas, nous allons disparaître », pronostiquait mi-août M. Vasquez. Les autorités panaméennes et les compagnies maritimes se refusent à évoquer le scénario catastrophe d’une possible interruption des activités du canal, si la sécheresse devait durer.

Reuters : Upheavals in Xi's world spread concern about China's diplomacy

Upheavals in Xi's world spread concern about China's diplomacy

BEIJING, Sept 17 (Reuters) - The disappearance of China's defence minister, the latest in a string of upheavals in the country's top ranks, is stoking uncertainty about President Xi Jinping's rule as an internal security clampdown trumps international engagement.

The growing unpredictability could affect the confidence other countries place in the leadership of the world's second-biggest economy, diplomats and analysts say.

Defence Minister Li Shangfu, who has missed meetings including with at least one foreign counterpart since he was last seen in late August, is under investigation in a corruption probe into military procurement, Reuters reported on Friday.

Newly installed Foreign Minister Qin Gang vanished with scant explanation in July, the same month as an abrupt shake-up of the military's elite Rocket Force, which oversees China's nuclear arsenal.

As Xi, China's commander-in-chief has focussed inward, he caused concern among foreign diplomats this month by missing a Group of 20 summit in India, the first time he has skipped the global leaders' gathering in his decade in power.

Faced with the growing uncertainties, some diplomats and analysts are calling for a hard look at the true nature of Xi's regime.

"Clear-eyed assessments are needed - this isn't just a question of whether China is a partner or a competitor, it is a source of economic, political and military risk," said Drew Thompson, a former Pentagon official who is now a scholar at the National University of Singapore.

Due to a lack of transparency surrounding the changes, various explanations were plausible "and this feeds the crisis of confidence that is brewing around China," Thompson said.

China's Foreign Ministry did not immediately respond to a request for comment on Saturday.

PROXIMITY ISN'T PATRONAGE
Regarding Defence Minister Li's disappearance and investigation, a ministry spokeswoman told reporters on Friday she was not aware of the situation. The State Council and Defence Ministry did not respond to requests for comment.

Since his appointment in March, Li has been the public face of China’s expanding military diplomacy, expressing concern over U.S. military operations during a high-profile security conference in June and visiting Russia and Belarus in August.

He had been expected to host an international security meeting in Beijing in October and represent the People's Liberation Army (PLA) at a meeting in November of regional defence chiefs in Jakarta.

With corruption long permeating China's military and state institutions, some analysts and diplomats believe Xi's anti-graft crackdowns mark political purges across the Communist Party.

"Regardless of the reason... the sense that this could keep happening could have an impact on foreign actors' confidence in engaging with their Chinese counterparts," said Helena Legarda, lead analyst with the Mercator Institute for China Studies in Berlin.

The Li upheaval is unusual for its speed and its reach into Xi's hand-picked elites.

"This is all so sudden and opaque. One thing we can now see is that proximity does not equate to patronage in Xi's world," said Singapore-based security analyst Alexander Neill, an adjunct fellow with Hawaii's Pacific Forum think-tank.

CONTINUITY RISK
Although not in a direct command position, Li serves on Xi's seven-person Central Military Commission and is one of China's five state councillors, a cabinet position that outranks regular ministers. Some scholars believe he is close to General Zhang Youxia, who sits above him on the commission and is Xi's closest ally in the PLA.

Li, sanctioned by Washington in 2018 for an arms deal with Russia, shunned a meeting with U.S. Defence Secretary Lloyd Austin at Singapore's Shangri-la Dialogue security conference in June, where a handshake marked their closest interaction.

Austin and other U.S. officials are keen to resume high-level talks between the two militaries regional tensions roil. But Beijing counters that it wants Washington to be less assertive in the Asia-Pacific.

Regional envoys say deeper Chinese military diplomacy is vital, particularly with the U.S. but also with other powers, as China increasingly deploys forces around Taiwan - the democratically governed island it claims - and across disputed parts of the East and South China Seas.

If Li's fate "reflects Xi's increasingly inward focus, it is not good for those of us who want greater openness and lines of communications with China's military," said one Asian diplomat.

As the PLA has an unprecedented level of military engagements with Southeast Asian forces this year, the recent swift changes back in Beijing "spur speculation and some concern about the continuity of policy", said political scientist Ja Ian Chong at the National University of Singapore.

"A shake-up of the military at this time is likely to draw attention, given the heightened activity of the PLA near Taiwan and the East China Sea, as well as stepped-up paramilitary activity in the South China Sea, since such actions create potential risk of accidents, escalation and crises," Chong said.

Reporting By Greg Torode in Hong Kong and Martin Quin Pollard in Beijing; Editing by William Mallard

Wired : The AI Detection Arms Race Is On

The AI Detection Arms Race Is On
And college students are developing the weapons, quickly building tools that identify AI-generated text—and tools to evade detection.

EDWARD TIAN DIDN’T think of himself as a writer. As a computer science major at Princeton, he’d taken a couple of journalism classes, where he learned the basics of reporting, and his sunny affect and tinkerer’s curiosity endeared him to his teachers and classmates. But he describes his writing style at the time as “pretty bad”—formulaic and clunky. One of his journalism professors said that Tian was good at “pattern recognition,” which was helpful when producing news copy. So Tian was surprised when, sophomore year, he managed to secure a spot in John McPhee’s exclusive non-fiction writing seminar.

Every week, 16 students gathered to hear the legendary New Yorker writer dissect his craft. McPhee assigned exercises that forced them to think rigorously about words: Describe a piece of modern art on campus, or prune the Gettysburg Address for length. Using a projector and slides, McPhee shared hand-drawn diagrams that illustrated different ways he structured his own essays: a straight line, a triangle, a spiral. Tian remembers McPhee saying he couldn’t tell his students how to write, but he could at least help them find their own unique voice.

If McPhee stoked a romantic view of language in Tian, computer science offered a different perspective: language as statistics. During the pandemic, he’d taken a year off to work at the BBC and intern at Bellingcat, an open source journalism project, where he’d written code to detect Twitter bots. As a junior, he’d taken classes on machine learning and natural language processing. And in the fall of 2022, he began to work on his senior thesis about detecting the differences between AI-generated and human-written text.

When ChatGPT debuted in November, Tian found himself in an unusual position. As the world lost its mind over this new, radically improved chatbot, Tian was already familiar with the underlying GPT-3 technology. And as a journalist who’d worked on rooting out disinformation campaigns, he understood the implications of AI-generated content for the industry.
While home in Toronto for winter break, Tian started playing around with a new program: a ChatGPT detector. He posted up at his favorite café, slamming jasmine tea, and stayed up late coding in his bedroom. His idea was simple. The software would scan a piece of text for two factors: “perplexity,” the randomness of word choice; and “burstiness,” the complexity or variation of sentences. Human writing tends to rate higher than AI writing on both metrics, which allowed Tian to guess how a piece of text had been created. Tian called the tool GPTZero—the “zero” signaled truth, a return to basics—and he put it online the evening of January 2. He posted a link on Twitter with a brief introduction. The goal was to combat “increasing AI plagiarism,” he wrote. “Are high school teachers going to want students using ChatGPT to write their history essays? Likely not.” Then he went to bed.

Tian woke up the next morning to hundreds of retweets and replies. There was so much traffic to the host server that many users couldn’t access it. “It was totally crazy,” Tian says. “My phone was blowing up.” A friend congratulated him on winning the internet. Teens on TikTok called him a narc. “A lot of the initial hate was like, ‘This kid is a snitch, he doesn’t have a life, he never had a girlfriend,’” says Tian with a grin. “Classic stuff.” (Tian has a girlfriend.) Within days, he was fielding calls from journalists around the world, eventually appearing on everything from NPR to the South China Morning Post to Anderson Cooper 360. Within a week, his original tweet had reached more than 7 million views.

GPTZero was a new twist in the media narrative surrounding ChatGPT, which had inspired industrywide hand-wringing and a scourge of AI-generated ledes. (Researchers had created a detector for GPT-2 text in 2019, but Tian’s was the first to target ChatGPT.) Teachers thanked Tian for his work, grateful they could finally prove their suspicions about fishy student essays. Had humanity found its savior from the robot takeover?

Tian’s program was a starting gun of sorts. The race was now on to create the definitive AI detection tool. In a world increasingly saturated with AI-generated content, the thinking went, we’ll need to distinguish the machine-made from the human-made. GPTZero represented a promise that it will indeed be possible to tell one from the other, and a conviction that the difference matters. During his media tour, Tian—smiley, earnest, the A student incarnate—elaborated on this reassuring view that no matter how sophisticated generative AI tools become, we will always be able to unmask them. There’s something irreducible about human writing, Tian said: “It has an element that can never be put into numbers.”

LIFE ON THE internet has always been a battle between fakers and detectors of fakes, with both sides profiting off the clash. Early spam filters sifted emails for keywords, blocking messages with phrases like “FREE!” or “be over 21,” and they eventually learned to filter out entire styles of writing. Spammers responded by surrounding their pitches with snippets of human-sounding language lifted from old books and mashed together. (This type of message, dubbed “litspam,” became a genre unto itself.) As search engines grew more popular, creators looking to boost their pages’ rankings resorted to “keyword stuffing”—repeating the same word over and over—to get priority. Search engines countered by down-ranking those sites. After Google introduced its PageRank algorithm, which favored websites with lots of inbound links, spammers created entire ecosystems of mutually supporting pages.

Around the turn of the millennium, the captcha tool arrived to sort humans from bots based on their ability to interpret images of distorted text. Once some bots could handle that, captcha added other detection methods that included parsing images of motorbikes and trains, as well as sensing mouse movement and other user behavior. (In a recent test, an early version of GPT-4 showed that it knew how to hire a person on Taskrabbit to complete a captcha on its behalf.) The fates of entire companies have rested on the issue of spotting fakes: Elon Musk, in an attempt to wriggle out of his deal to buy Twitter, cited a bot detector to boost his argument that Twitter had misrepresented the number of bots on its site.

Generative AI re-upped the ante. While large language models and text-to-image generators have been evolving steadily over the past decade, 2022 saw an explosion of consumer-friendly tools like ChatGPT and Dall-E. Pessimists argue that we could soon drown in a tsunami of synthetic media. “In a few years, the vast majority of the photos, videos, and text we encounter on the internet could be AI-generated,” New York Times technology columnist Kevin Roose warned last year. The Atlantic imagined a looming “textpocalypse” as we struggle to filter out the generative noise. Political campaigns are leveraging AI tools to create ads, while Amazon is flooded with ChatGPT-written books (many of them about AI). Scrolling through product reviews already feels like the world’s most annoying Turing test. The next step seems clear: If you thought Nigerian prince emails were bad, wait until you see Nigerian prince chatbots.

Soon after Tian released GPTZero, a wave of similar products appeared. OpenAI rolled out its own detection tool at the end of January, while Turnitin, the anti-plagiarism giant, unveiled a classifier in April. They all shared a basic methodology, but each model was trained on different data sets. (For example, Turnitin focused on student writing.) As a result, precision varied wildly, from OpenAI’s claim of 26 percent accuracy for detecting AI-written text, up to the most optimistic claim from a company called Winston AI at 99.6 percent. To stay ahead of the competition, Tian would have to keep improving GPTZero, come up with its next product, and finish college in the meantime.

Right away, Tian recruited his high school friend Alex Cui as CTO and, over the following weeks, brought on a handful of programmers from Princeton and Canada. Then, in the spring, he onboarded a trio of coders from Uganda, whom he’d met four years earlier while working for a startup that trains engineers in Africa. (A global citizen, Tian was born in Tokyo and lived in Beijing until age 4 before his parents, both Chinese engineers, moved the family to Ontario.) Together the team began working on its next application: a Chrome plug-in that would scan the text of a web page and determine whether it was AI-generated.

Another threat to GPTZero was GPTZero. Almost immediately after it launched, skeptics on social media started posting embarrassing examples of the tool misclassifying texts. Someone pointed out that it flagged portions of the US Constitution as possibly AI-written. Mockery gave way to outrage when stories of students falsely accused of cheating due to GPTZero began to flood Reddit. At one point, a parent of one such student reached out to Soheil Feizi, a professor of computer science at the University of Maryland. “They were really furious,” Feizi said. Last fall, before GPTZero debuted, Feizi and some other Maryland colleagues had begun putting together a research project on the problems with AI detectors, which he’d suspected might not be reliable. Now GPTZero and its imitators got him thinking they could do more harm than good.

Yet another headache for Tian was the number of crafty students finding ways around the detector. One person on Twitter instructed users to insert a zero-width space before every “e” in a ChatGPT-generated text. A TikTok user wrote a program that bypassed detection by replacing certain English letters with their Cyrillic look-alikes. Others started running their AI text through QuillBot, a popular paraphrasing tool. Tian patched these holes, but the workarounds kept coming. It was only a matter of time before someone spun up a rival product—an anti-detector.

IN EARLY MARCH, a Stanford University freshman named Joseph Semrai and some friends were driving down the Pacific Coast Highway to LA when they got locked out of their Zipcar in Ventura. They walked to a nearby Starbucks and waited for roadside assistance. But as the wait dragged on for hours, Semrai and a friend wondered how to make up for the lost time. Semrai had an essay due the following week for a required freshman writing class. It was his least favorite type of assignment: a formulaic essay meant to show logical reasoning. “It’s a pretty algorithmic process,” says Semrai.

ChatGPT was the obvious solution. But at the time, its responses tended to max out at a few paragraphs, so generating a full-length essay would be a multistep process. Semrai wanted to create a tool that could write the paper in one burst. He also knew there was a chance it could be detected by GPTZero. With the encouragement of his friend, Semrai pulled out his laptop and ginned up a script that would write an essay based on a prompt, run the text through GPTZero, then keep tweaking the phrasing until the AI was no longer detectable—essentially using GPTZero against itself.

Semrai introduced his program a few days later at Friends and Family Demo Day, a kind of show-and-tell for Stanford’s undergraduate developer community. Standing before a roomful of classmates, he asked the audience for an essay topic—someone suggested “fine dining” in California—and typed it into the prompt box. After a few seconds, the program spat out an eight-paragraph essay, unoriginal but coherent, with works cited. “Not saying I’d ever submit this paper,” Semrai said, to chuckles. “But there you go. I dunno, it saves time.” He named the tool WorkNinja and put it on the app store two months later. With the help of a promotional campaign featuring the Gen Z influencer David Dobrik and a giveaway of 10 Teslas to users who signed up, it received more than 350,000 downloads in the first week; sign-ups have slowed since then to a few hundred a day, according to Semrai. (Semrai wouldn’t say who funded the campaign, only that it was a major Silicon Valley angel investor.)

Semrai’s Zoomer mop and calm demeanor belie a simmering intensity. Whereas Tian bounces and bubbles his way through the world, Semrai comes across as focused and deadpan. The 19-year-old speaks in the confident, podcast-ready tone of a Silicon Valley entrepreneur who sees the world in terms of problems to be solved, ending every other sentence with, “Right?” Listening to him wax on about defensible moats and the “S-curve” of societal growth, it’s easy to forget he can’t legally drink. But then, occasionally, he’ll say something that reveals the wide-eyed undergrad, open to the world and still figuring out his place in it. Like the time he and a friend walked around the Santa Monica pier until 3 am, “talking about what we value.” Semrai thinks a lot about how to find balance and happiness. “I think, while I’m young, it probably lies more in exploring the derivative,” he says, “chasing the highs and lows.”

Growing up in New York and then Florida, his parents—a firefighter father from Yonkers and a homemaker mother from China—gave him a long leash. “I was kinda left during childhood to pursue what genuinely excited me,” he said. “The best way to do that was to make stuff on the computer.” When Semrai was 6 he created a plug-in to assign permission levels for Minecraft servers, and at 7 he wrote a program that patched Windows 7 so you could run Windows XP on it. “It just makes me genuinely happy to ship things for people,” he says.

His family moved from Queens to Palm City when he was 9, and Semrai saw the difference between the public school systems. The basic computer literacy he’d taken for granted in New York schools was scarce in Florida. He started writing programs to help fill gaps in education—a trajectory that allows him to say, at 19, that he’s been “working in ed tech my entire life.” Freshman year of high school, he created an online learning platform that won startup funding in a local competition. Prior to Covid, he’d created a digital hall pass system that became the basis for contact tracing and was adopted by 40 school districts in the Southeast.

Semrai is fundamentally a techno-optimist. He says he believes that we should speed the development of technology, including artificial general intelligence, because it will ultimately lead us toward a “post-scarcity” society—a worldview sometimes described as “effective accelerationism.” (Not to be confused with effective altruism, which holds that we should take actions that maximize “good” outcomes, however defined.) Semrai’s case for WorkNinja rests on its own kind of accelerationist logic. AI writing tools are good, in his view, not because they help kids cheat, but because they’ll force schools to revamp their curricula. “If you can follow a formula to create an essay, it’s probably not a good assignment,” he says. He envisions a future in which every student can get the kind of education once reserved for aristocrats, by way of personalized AI tutoring. When he was first learning how to program, Semrai says, he relied largely on YouTube videos and internet forums to answer his questions. “It would have been easier if there was a tutor to guide me,” he says. Now that AI tutors are real, why stand in their way?
Joseph Semrai, 19, creator of WorkNinja, a tool that generates AI-written essays.
ILLUSTRATION: JAMES MARSHALL; PHOTOGRAPH: CHARIS MORGAN

I RECENTLY USED WorkNinja to generate a handful of essays, including one about Darwin’s theory of evolution. The first version it gave me was clumsy and repetitive, but workable, exploring the theory’s implications for biology, genetics, and philosophy. GPTZero flagged it as likely AI-generated.

So I hit WorkNinja’s Rephrase button. The text shifted slightly, replacing certain words with synonyms. After three rephrasings, GPTZero finally gave the text its stamp of humanity. (When I tested the same text again a few weeks later, the tool labeled it a mix of human and AI writing.) The problem was, many of the rephrased sentences no longer made sense. For example, the following sentence:
Darwin’s theory of evolution is the idea that living species evolve over time due to their interaction with their environment.

had morphed to become:
Darwin’s theory of evolution is the thought that living species acquire over clip due to their interaction with their surroundings.
At the very least, any student looking for a shortcut would have to clean up their WorkNinja draft before submitting it. But it points to a real issue: If even this janky work in progress can circumvent detectors, what could a sturdier product accomplish?
In March, Soheil Feizi at the University of Maryland published his findings on the performance of AI detectors. He argued that accuracy problems are inevitable, given the way AI text detectors worked. As you increase the sensitivity of the instrument to catch more AI-generated text, you can’t avoid raising the number of false positives to what he considers an unacceptable level. So far, he says, it’s impossible to get one without the other. And as the statistical distribution of words in AI-generated text edges closer to that of humans—that is, as it becomes more convincing—he says the detectors will only become less accurate. He also found that paraphrasing baffles AI detectors, rendering their judgments “almost random.” “I don’t think the future is bright for these detectors,” Feizi says.

“Watermarking” doesn’t help either, he says. Under this approach, a generative AI tool like ChatGPT proactively adjusts the statistical weights of certain interchangeable “token” words—say, using start instead of begin, or pick instead of choose—in a way that would be imperceptible to the reader but easily spottable by an algorithm. Any text in which those words appear with a given frequency could be marked as having been generated by a particular tool. But Feizi argues that with enough paraphrasing, a watermark “can be washed away.”

In the meantime, he says, detectors are hurting students. Say a detection tool has a 1 percent false positive rate—an optimistic assumption. That means in a classroom of 100 students, over the course of 10 take-home essays, there will be on average 10 students falsely accused of cheating. (Feizi says a rate of one in 1,000 would be acceptable.) “It’s ridiculous to even think about using such tools to police the use of AI models,” he says.

Tian says the point of GPTZero isn’t to catch cheaters, but that has inarguably been its main use case so far. (GPTZero’s detection results now come with a warning: “These results should not be used to punish students.”) As for accuracy, Tian says GPTZero’s current level is 96 percent when trained on its most recent data set. Other detectors boast higher figures, but Tian says those claims are a red flag, as it means they’re “overfitting” their training data to match the strengths of their tools. “You have to put the AI and human on equal footing,” he says.

Surprisingly, AI-generated images, videos, and audio snippets are far easier to detect, at least for now, than synthetic text. Reality Defender, a startup backed by Y Combinator, launched in 2018 with a focus on fake image and video detection and has since branched out to audio and text. Intel released a tool called FakeCatcher, which detects deepfake videos by analyzing facial blood flow patterns visible only to the camera. A company called Pindrop uses voice “biometrics” to detect spoofed audio and to authenticate callers in lieu of security questions.

AI-generated text is more difficult to detect because it has relatively few data points to analyze, which means fewer opportunities for AI output to deviate from the human norm. Compare that to Intel’s FakeCatcher. Ilke Demir, a research scientist for Intel who has also worked on Pixar films, says it would be extremely difficult to create a data set large and detailed enough to allow deepfakers to simulate blood flow signatures to fool the detector. When I asked whether such a thing could eventually be created, she said her team anticipates future developments in deepfake technology in order to stay ahead of them.

Ben Colman, CEO of Reality Defender, says his company’s detection tools are unevadable in part because they’re private. (So far, the company’s clients have mainly been governments and large corporations.) With publicly available tools like GPTZero, anyone can run a piece of text through the detector and then tweak it until it passes muster. Reality Defender, by contrast, vets every person and institution that uses the tool, Colman says. They also watch out for suspicious usage, so if a particular account were to run tests on the same image over and over with the goal of bypassing detection, their system would flag it.

Regardless, much like spam hunters, spies, vaccine makers, chess cheaters, weapons designers, and the entire cybersecurity industry, AI detectors across all media will have to constantly adapt to new evasion techniques. Assuming, that is, the difference between human and machine still matters.

THE MORE TIME I spent talking with Tian and Semrai and their classmate-colleagues, the more I wondered: Do any of these young people actually … enjoy writing? “Yeah, a lot!” said Tian, beaming even more than usual when I asked him last May on Princeton’s campus. “It’s like a puzzle.” He likes figuring out how words fit together and then arranging the ideas so they flow. “I feel like that’s fun to do.” He also loves the interview process, as it gives him “a window into people’s lives, plus a mirror into how you live your own.”

In high school, Tian says, writing felt like a chore. He credits McPhee for stoking his love and expanding his taste. In June, he told me excitedly that he had just picked up a used copy of Annie Dillard’s The Writing Life.

Semrai similarly found high school writing assignments boring and mechanistic—more about synthesizing information than making something new. “I’d have preferred open-format assignments that would’ve sparked creativity,” he says. But he put those synthesizing skills to work. Sophomore year, he wrote an 800-page instructional book called Build for Anything, intended “to take someone from knowing nothing to knowing a little bit of almost everything” about web development. (He self-published the book on Amazon in 2022 and sold a few hundred copies.) Semrai said it’s the kind of prose ChatGPT now excels at. “I don’t think the book falls into the category of meaningful writing,” he says.

After almost 20 years of typing words for money, I can say from experience, writing sucks. Ask any professional writer and they’ll tell you, it’s the worst, and it doesn’t get easier with practice. I can attest that the enthusiasm and curiosity required to perpetually scan the world, dig up facts, and wring them for meaning can be hard to sustain. And that’s before you factor in the state of the industry: dwindling rates, shrinking page counts, and shortening attention spans (readers’ and my own). I keep it up because, for better or worse, it’s now who I am. I do it not for pleasure but because it feels meaningful—to me at least.
Some writers romanticize the struggle. McPhee once described lying on a picnic table for two weeks, trying to decide how to start an article. “The piece would ultimately consist of some five thousand sentences, but for those two weeks I couldn’t write even one,” he wrote. Another time, at age 22, he lashed himself to his writing chair with a bathrobe belt. According to Thomas Mann, “A writer is someone for whom writing is more difficult than it is for other people.” “You search, you break your heart, your back, your brain, and then—only then—it is handed to you,” writes Annie Dillard in The Writing Life. She offers this after a long comparison of writing to alligator wrestling.

The implication is that the harder the squeeze, the sweeter the juice—that there’s virtue in staring down the empty page, taming it, forcing it to give way to prose. This is how the greatest breakthroughs happen, we tell ourselves. The agony is worth it, because that’s how ideas are born.

The siren call of AI says, It doesn’t have to be this way. And when you consider the billions of people who sit outside the elite club of writer-sufferers, you start to think: Maybe it shouldn’t be this way.

MAY HABIB SPENT her early childhood in Lebanon before moving to Canada, where she learned English as a second language. “I thought it was pretty unfair that so much benefit would accrue to someone really good at reading and writing,” she says. In 2020, she founded Writer, one of several hybrid platforms that aims not to replace human writing, but to help people—and, more accurately, brands—collaborate better with AI.

Habib says she believes there’s value in the blank page stare-down. It helps you consider and discard ideas and forces you to organize your thoughts. “There are so many benefits to going through the meandering, head-busting, wanna-kill-yourself staring at your cursor,” she says. “But that has to be weighed against the speed of milliseconds.”

The purpose of Writer isn’t to write for you, she says, but rather to make your writing faster, stronger, and more consistent. That could mean suggesting edits to prose and structure, or highlighting what else has been written on the subject and offering counterarguments. The goal, she says, is to help users focus less on sentence-level mechanics and more on the ideas they’re trying to communicate. Ideally, this process yields a piece of text that’s just as “human” as if the person had written it entirely themselves. “If the detector can flag it as AI writing, then you’ve used the tools wrong,” she says.

The black-and-white notion that writing is either human- or AI-generated is already slipping away, says Ethan Mollick, a professor at the Wharton School of the University of Pennsylvania. Instead, we’re entering an era of what he calls “centaur writing.” Sure, asking ChatGPT to spit out an essay about the history of the Mongol Empire produces predictably “AI-ish” results, he says. But “start writing, ‘The details in paragraph three aren’t quite right—add this information, and make the tone more like The New Yorker,’” he says. “Then it becomes more of a hybrid work and much better-quality writing.”

Mollick, who teaches entrepreneurship at Wharton, not only allows his students to use AI tools—he requires it. “Now my syllabus says you have to do at least one impossible thing,” he says. If a student can’t code, maybe they write a working program. If they’ve never done design work, they might put together a visual prototype. “Every paper you turn in has to be critiqued by at least four famous entrepreneurs you simulate,” he says.

Students still have to master their subject area to get good results, according to Mollick. The goal is to get them thinking critically and creatively: “I don’t care what tool they’re using to do it, as long as they’re using the tools in a sophisticated manner and using their mind.”

Mollick acknowledges that ChatGPT isn’t as good as the best human writers. But it can give everyone else a leg up. “If you were a bottom-quartile writer, you’re in the 60th to 70th percentile now,” he says. It also frees certain types of thinkers from the tyranny of the writing process. “We equate writing ability with intelligence, but that’s not always true,” he says. “In fact, I’d say it’s often not true.”


Edward Tian, age 23, creator of GPTZero, a tool that detects AI-generated writing.
ILLUSTRATION: JAMES MARSHALL; PHOTOGRAPH: LAURYN HILL

ON A CLOUDLESS day in May, Tian and I strolled across Princeton’s campus; big white reunion tents seemed to have landed like spaceships on the manicured lawns. At my request, Tian invited a handful of classmates to join us for lunch at a Szechuan restaurant off campus and talk about AI.

As some schools rushed to ban ChatGPT and tech CEOs signed letters warning of AI-fueled doom, the students were notably relaxed about a machine-assisted future. (Princeton left it up to professors to set their own ground rules.) One had recently used ChatGPT to write the acknowledgments section of her thesis. Others, including Tian, relied on it to fill in chunks of script while coding. Lydia You, a senior and computer science major who plans to work in journalism, had asked ChatGPT to write a poem about losing things in the style of Elizabeth Bishop—an attempt to re-create her famous poem “One Art.” (“The art of losing isn’t hard to master.”) The result was “very close” to the original poem, You said, and she found that the chatbot did an even better job analyzing the original and describing what made it so moving. “We’ve seen a lot of panic about almost everything in our lives,” said You, citing TikTok, Twitter, and the internet itself. “I feel like people of our generation are like, We can figure out for ourselves how to use this.”

Sophie Amiton, a senior studying mechanical and aerospace engineering, jumped in: “Also, I think our generation is lazier in a lot of ways,” she said, as You nodded in agreement. “I see a lot more people who don’t want traditional jobs now, don’t want a nine-to-five.”

“They’re disillusioned,” You said. “A lot of jobs are spreadsheets.”

“I think that came out of Covid,” Amiton continued. “People reevaluated what the purpose of work even is, and if you can use ChatGPT to make your life easier, and therefore have a better quality of life or work-life balance, then why not use the shortcut?”

Liz, a recent Princeton graduate who preferred not to use her surname, sent me a paper she’d written with the help of ChatGPT for a class on global politics. Rather than simply asking it to answer the essay question, she plugged in an outline with detailed bullet points, then had it write the paper based on her notes. After extensive back-and-forth—telling it to rewrite and rearrange, add nuance here and context there—she finally had a paper she was comfortable submitting. She got an A.

I copied and pasted her paper into GPTZero. The verdict: “Your text is likely to be written entirely by a human.”
In early May, just a few weeks before Tian and his classmates put on their black graduation gowns, the GPTZero team released the Chrome plug-in they’d been developing and called it Origin. Origin is still rudimentary: You have to select the text of a web page yourself, and its accuracy isn’t perfect. But Tian hopes that one day the tool will automatically scan every website you look at, highlighting AI-generated content—from text to images to video—as well as anything “toxic” or factually dubious. He describes Origin as a “windshield” for the information superhighway, deflecting useless or harmful material and allowing us to see the road clearly.

Tian was unflaggingly optimistic about the company; he also just felt fortunate to be graduating into a job he actually wanted. Many of his friends had entered Princeton planning to be entrepreneurs, but belt-tightening in the tech sector had changed their plans.
As a rising sophomore with three years left to go at Stanford, Semrai approached the summer with a more footloose attitude. On a blistering Thursday afternoon in June, on the rooftop of Pier 17 near Wall Street, Semrai, wearing a green patterned shirt and white Nikes, spoke to me brightly about the future—or at least the next few weeks. His summer was still taking shape. (“I’m rapidly thesis-testing.”) But for now he was in New York, crashing with friends while cranking on a couple of AI-driven projects. The previous night, he’d slept in a coworking space in SoHo. Now he was standing in the shaded VIP section of an event put on by Techstars New York City, a startup accelerator, while hundreds of sweat-stained attendees milled around in the glare.

Nearby, New York City mayor Eric Adams stood onstage wearing aviators and a full suit, praising the glories of coding. “I’m a techie,” Adams said, before encouraging guests to seek out diverse collaborators and use “source code” to fix societal problems like cancer and gun violence. He then urged the singles in the crowd to find themselves a “shorty or a boo and hook up with them.”

Semrai was taking a see-what-sticks approach to building. In addition to WorkNinja, he was developing a platform for chatbots based on real celebrities and trained on reams of their data, with which fans could then interact. He was also prototyping a bracelet that would record everything we say and do—Semrai calls it a “perfect memory”—and offer real-time tips to facilitate conversations. (A group of classmates at Stanford recently created a related product called RizzGPT, an eyepiece that helps its wearer flirt.)

He expected the summer to give rise to an explosion of AI apps, as young coders mix and cross-pollinate. (Eric Adams would approve.) “I think a constellation of startups will be formed, and five years from now we’ll be able to draw lines between people—the start of an ecosystem,” he said.

By summer, Tian had a team of 12 employees and had raised $3.5 million from a handful of VCs, including Jack Altman (brother of OpenAI CEO Sam Altman) and Emad Mostaque of Stability AI. But over the course of our conversations, I noticed that his framing of GPTZero/Origin was shifting slightly. Now, he said, AI-detection would be only one part of the humanity-proving toolkit. Just as important would be an emphasis on provenance, or “content credentials.” The idea is to attach a cryptographic tag to a piece of content that verifies it was created by a human, as determined by its process of creation—a sort of captcha for digital files. Adobe Photoshop already attaches a tag to photos that harness its new AI generation tool, Firefly. Anyone looking at an image can right-click it and see who made it, where, and how. Tian says he wants to do the same thing for text and that he has been talking to the Content Authenticity Initiative—a consortium dedicated to creating a provenance standard across media—as well as Microsoft about working together.

One could interpret his emphasis on provenance as a tacit acknowledgment that detection alone won’t cut it. (OpenAI shut down its text classifier in July “due to its low rate of accuracy.”) It also previews a possible paradigm shift in how we relate to digital media. The whole endeavor of detection suggests that humans leave an unmistakable signature in a piece of text—something perceptible—much the way that a lie detector presumes dishonesty leaves an objective trace. Provenance relies on something more like a “Made in America” label. If it weren’t for the label, we wouldn’t know the difference. It’s a subtle but meaningful distinction: Human writing may not be better, or more creative, or even more original. But it will be human, which will matter to other humans.

In June, Tian’s team took another step in the direction of practicality. He told me they were building a new writing platform called HumanPrint, which would help users improve their AI-written text and enable them to share “proof of authenticity.” Not by generating text, though. Rather, it would use GPTZero’s technology to highlight sections of text that were insufficiently human and prompt the user to rewrite it in their own words—a sort of inversion of the current AI writing assistants. “So teachers can specify, OK, maybe more than 50 percent of the essay should still be written in your own words,” he said. When I asked whether this was a pivot for the company, Tian argued that it was “a natural extension of detection.” “It was always a vision of being the gold standard of responsible AI usage,” Tian said, “and that’s still there.” Still, the implication is clear: There’s no stopping AI writing; the only option is to work with it.

WHEN TIAN WAS first testing out GPTZero, he scanned a 2015 New Yorker essay by McPhee called “Frame of Reference.” In it, McPhee riffs on the joys and risks of making cultural references in one’s writing. “Mention Beyoncé and everyone knows who she is. Mention Veronica Lake and you might as well be in the Quetico-Superior,” he writes coyly. He runs down a list of adjectives he’s used to describe mustaches, including “sincere,” “no-nonsense,” “gyroscopic,” “guileless,” “analgesic,” “soothing,” “odobene,” and “tetragrammatonic.” He concludes with an anecdote about battling an editor to include a reference to an obscure British term used by upper-class tourists to India during the Raj. (He won.) It’s classic McPhee: scalpel-precise, big-hearted if a tad self-satisfied, gleefully digressive, indulgent until he gets to the just-right point. GPTZero determined that the article was “the most human on all metrics,” Tian said. I called McPhee to ask what he thought it meant that his writing was especially human.

“I really have no very good idea,” McPhee told me over the phone. “But if I were guessing, it’s that my pieces get at the science, or the agriculture, or the aviation, or whatever the topic is, through people. There’s always a central figure I learn from.” Indeed, McPhee writes through the eyes of experts. The reader comes away with not just some esoteric knowledge about geology or particle physics or oranges, but a sense of the person studying the subject, as well as McPhee studying the person.

McPhee, now 92 , said he’s unconcerned about AI replacing human writers. “I’m extremely skeptical and not the least bit worried about it,” he said. “I don’t think there’s a Mark Twain of artificial intelligence.”

But, I asked, what if years from now, someone designs a McPheeBot3000 trained on McPhee’s writing, and then asks it to produce a book on a fresh topic? It might not be able to ford streams with environmental activists or go fly-fishing with ichthyologists, but couldn’t it capture McPhee’s voice and style and worldview? Tian argued that machines can only imitate, while McPhee never repeats himself: “What’s unique about McPhee is he comes up with things McPhee a day ago wouldn’t have.”

I asked McPhee about the hypothetical McPheeBot3000. (Or, if Semrai has his way, not-so-hypothetical.) “If this thing ever happens, in a future where I’m no longer here,” he said, “I hope my daughters show up with a lawyer.”