El Partido Popular ha elevado el tono sobre lo ocurrido en Ceuta hasta hacer suyo un término que durante años parecía reservado a Vox: «invasión». La dirección popular se desgañita por justificar esa palabra, convencida de que el debate migratorio se ha convertido en un terreno decisivo para movilizar a su electorado. La elección no es neutral. Se trata de un marco que activa emociones, desplaza la discusión sobre derechos y fronteras hacia el terreno de la seguridad y obliga al resto de actores a posicionarse. El PP lo utiliza siempre que considera que le renta, aunque en el pasado lo hubiera evitado.

Alberto Núñez Feijóo ha defendido que calificar de invasión el episodio no es una exageración, sino una forma de describir una presión migratoria que, a su juicio, el Gobierno no controla. Santiago Abascal ha ido más lejos y ha presentado a Ceuta como la prueba de una supuesta estrategia socialista para diluir fronteras. Ambos coinciden en señalar a Pedro Sánchez como responsable último de una situación que mezcla diplomacia, cooperación con Marruecos, acogida y control fronterizo. El lenguaje, por tanto, no solo describe: también ordena el campo de batalla político.

Feijóo y Abascal acusan a Sánchez de manipular el censo con la ley de nietos

En paralelo al pulso sobre Ceuta, Feijóo y Abascal han acusado al Gobierno de manipular el censo con la ley de nietos. Sostienen que la norma de memoria democrática está sirviendo para incorporar de forma acelerada a nuevos votantes que, en su relato, engrosarían las filas del electorado afín a la izquierda. La acusación es grave porque pone en cuestión la limpieza del sistema electoral y la independencia de los organismos que gestionan el censo. No se presenta como una simple discrepancia política, sino como una denuncia de manipulación institucional.

Pedro Sánchez ha respondido con dureza: les ha reprochado que quieran «cercenar» el derecho al voto. El presidente ha enmarcado la ofensiva de PP y Vox en una estrategia más amplia de restricción de derechos, que comenzaría con los migrantes, seguiría con los descendientes de exiliados y terminaría afectando a los residentes en el exterior. La discusión, por tanto, ya no se limita a la ley de nietos: alcanza al modelo de ciudadanía, a la relación con la diáspora y a la propia arquitectura del censo electoral.

Qué es la ley de nietos y por qué está en el centro del debate

La conocida como ley de nietos es la disposición adicional octava de la Ley de Memoria Democrática, aprobada en 2022. Su objetivo es reparar la exclusión de quienes perdieron o renunciaron a la nacionalidad española por motivos políticos, ideológicos o de exilio. La norma permite que determinados hijos y nietos de españoles puedan solicitar la nacionalidad, amplía plazos y elimina requisitos que antes obligaban a renunciar a otra nacionalidad. No es, en sentido estricto, una ley electoral, pero su aplicación tiene efectos sobre el padrón, la nacionalidad y, en última instancia, sobre el censo.

• Reconoce el derecho a la nacionalidad para descendientes de exiliados y represaliados.
• Amplía el plazo de solicitud y flexibiliza las condiciones respecto a normas anteriores.
• Ha generado decenas de miles de expedientes, con especial incidencia en América Latina.
• Sus críticos sostienen que el Gobierno la utiliza para inflar el censo con nuevos votantes.
• Sus defensores subrayan que se trata de una reparación histórica y no de una operación partidista.

El punto central es que la nacionalidad no se concede de forma automática ni discrecional por el Gobierno. Los solicitantes deben acreditar vínculos, documentación y cumplir requisitos legales. Una vez obtenida la nacionalidad, la inscripción en el censo

Source: MSN News

Die Szenerie entlang des berühmten Pacific Coast Highway glich in den vergangenen Tagen einem Katastrophengebiet. Einsatzkräfte, Statiker und Sicherheitspersonal waren im Dauereinsatz, um beschädigte Gebäude zu prüfen und weitere Unglücke zu verhindern. Besonders betroffen sind Grundstücke, die direkt an der Steilküste liegen. Dort treffen die gewaltigen Wellen auf Böschungen, die durch Regenfälle zusätzlich aufgeweicht wurden. Bei Nicolas Cage führte diese Kombination dazu, dass die unterirdischen Stützpfeiler der Auffahrt komplett ausgewaschen wurden.

Die wichtigsten Fakten im Überblick

• Betroffen ist das Malibu-Anwesen von Schauspieler Nicolas Cage.
• Hurrikan Marie und starke Brandung haben die kalifornische Küste schwer getroffen.
• Teile der Einfahrt sind in einem Krater eingebrochen; Wasser bewegt sich zwischen freigelegten Stützpfeilern.
• Das Haus hat vier Stockwerke und rund 366 Quadratmeter Wohnfläche.
• Ob saniert oder abgerissen wird, ist noch unklar.
• Der Schauspieler war möglicherweise nicht vor Ort; genaue Umstände sind nicht bestätigt.
• Entlang des Pacific Coast Highway waren zahlreiche Einsatzkräfte im Einsatz.

Die Bilder aus Malibu zeigen, wie tief die Zerstörung reicht. Ganze Teile des Hauses mit Traumausblick auf den Ozean wurden beschädigt oder sind weggebrochen. Der Krater in der Auffahrt ist dabei nur das sichtbarste Symptom eines Problems, das sich unter der Oberfläche abspielt. Das eindringende Meerwasser hat den Boden unter dem Anwesen instabil gemacht. Fachleute warnen seit Jahren, dass die Kombination aus steigendem Meeresspiegel, stärkeren Stürmen und intensiveren Regenfällen die Küstenregionen Südkaliforniens verwundbarer macht.

Was in Malibu geschah

Die jüngsten Schäden gehen auf Hurrikan Marie und die damit verbundene starke Brandung zurück. Zwar erreichte der Sturm die Küste nicht als direkter Wirbelsturm, doch seine Ausläufer peitschten monströse Wellen gegen die Klippen. Diese Wellen trafen auf Böden, die durch heftige Regenfälle bereits durchweicht waren. In Malibu führte das zu Unterspülungen, Erdrutschen und eingestürzten Auffahrten. Zahlreiche Liegenschaften der meist gut betuchten Anrainerschaft wurden in Mitleidenschaft gezogen.

Ein Anwesen sticht besonders hervor: die Immobilie von Nicolas Cage. Der Schauspieler, der 1964 geboren wurde und zu den bekanntesten Gesichtern Hollywoods zählt, besitzt seit Jahren mehrere Immobilien. Seine Malibu-Villa gilt als Rückzugsort mit spektakulärem Meerblick. Nun ist genau dieser Standort zum Problem geworden. Das vierstöckige Gebäude mit rund 366 Quadratmetern Wohnfläche steht auf einem Grundstück, das von den Naturgewalten unmittelbar bedroht ist.

Ob Cage zum Zeitpunkt des Unglücks anwesend war, ist nicht bekannt. Die Behörden haben den Bereich nach Berichten aus US-Medien abgesperrt. Fachleute prüfen, ob das Haus noch standsicher ist. In der Auffahrt ist ein großer Krater entstanden, in dem sich Wasser sammelt. Zwischen den freigelegten Stützpfeilern bewegt sich die See, was die gesamte Konstruktion weiter schwächt. Ein solcher Schaden ist nicht nur kosmetisch, sondern kann die Statik des gesamten Anwesens gefährden.

Nicolas Cage und sein Verhältnis zu Immobilien

Nicolas Cage ist nicht nur für seine Filmrollen bekannt, sondern auch für seine exzentrischen Immobilienkäufe. Im Laufe seiner Karriere erwarb er Schlösser, Inseln, Villen und historische Anwesen. Viele davon verkaufte er später wieder, teils mit Verlust. Sein Malibu-Haus gehört zu den Objekten, die er über die Jahre behielt oder die immer wieder in den Mittelpunkt öffentlicher Berichterstattung gerieten. Der Schauspieler hat in Interviews mehrfach über seine finanzielle Situation gesprochen und erklärt, dass er aus Fehlern gelernt habe.

Für Cage dürfte die Versicherungsfrage nun eine zentrale Rolle spielen. Küstenschäden sind in Kalifornien ein heikles Thema. Viele Versicherer schließen Überschwemmungs- und Erdrutschschäden aus oder verlangen hohe Prämien. Ob seine Police die aktuellen Zerstörungen abdeckt, ist öffentlich nicht bekannt. Selbst wenn nicht, hätte der Schauspieler nach Einschätzung von Branchenkennern genügend Mittel, um eine Sanierung oder einen Neubau zu finanzieren. Allerdings ist Geld nicht das einzige Problem: Wenn die Küste weiter erodiert, könnte ein Wiederaufbau am selben Ort langfristig riskant sein.

Warum die Küste so verwundbar ist

Malibu ist ein Symbol für den kalifornischen Traum: Sonne, Meer, Palmen und spektakuläre Häuser an steilen Hängen. Doch genau diese Lage macht die Region anfällig. Der Pazifik nagt permanent an den Klippen. Bei Stürmen und Hurrikanen werden die Wellen zu einer Waffe, die Felsen und Böden innerhalb von Stunden wegreißen kann. Der Klimawandel verstärkt dieses Risiko. Wärmere Ozeane liefern mehr Energie für Stürme, während steigende Pegel die Küstenlinie nach innen drücken.

Hinzu kommen menschliche Eingriffe. Uferbefestigungen, Mauern und Entwässerungssysteme können kurzfristig schützen, aber sie verändern auch die Strömung und können Erosion an anderen Stellen verstärken. In Malibu gibt es seit Jahren Diskussionen über den Umgang mit gefährdeten Grundstücken. Einige Eigentümer investieren Millionen in Schutzbauten, andere klagen gegen Bauvorschriften. Die aktuellen Schäden dürften diese Debatte weiter anheizen.

Die Schäden im Detail

Der Krater in der Auffahrt von Cages Anwesen ist besonders auffällig. In dem Loch ist Wasser zu sehen, das sich zwischen den ehemals unterirdischen Stützpfeilern hin und her bewegt. Diese Pfeiler trugen zuvor den Boden und die Zufahrt. Jetzt sind sie freigelegt, korrodiert und teilweise unterspült. Das gesamte Konstrukt wurde ausgewaschen. Für Bauingenieure ist das ein Alarmsignal: Wenn die Fundamente nicht mehr stabil sind, kann ein Gebäude selbst dann einsturzgefährdet sein, wenn die Fassade noch intakt wirkt.

Teile des Hauses wurden beschädigt oder sind weggebrochen. Der Traumausblick auf die wilde Gischt des Pazifiks ist für Cage vorerst zum Albtraum geworden. Die Immobilie ist zur Großbaustelle mutiert. Baufahrzeuge, Absperrungen und Sicherheitskräfte prägen das Bild. Ob nur eine Sanierung oder ein Komplettabriss droht, steht zum jetzigen Zeitpunkt noch nicht fest. Diese Entscheidung hängt von mehreren Faktoren ab: der Standsicherheit, dem Zustand der Fundamente, den Auflagen der Behörden und nicht zuletzt von der Versicherung.

Wie es weitergehen könnte

Sollte das Haus abgerissen werden müssen, wäre das für Cage ein herber Verlust. Ein Neubau an derselben Stelle könnte Jahre dauern und Millionen kosten. Genehmigungen in Malibu sind streng, besonders in Küstennähe. Umweltschutzauflagen, Denkmalschutz und Nachbarschaftsinteressen können Projekte verzögern. Andererseits könnte ein Neubau die Chance bieten, das Anwesen besser gegen Stürme und Erosion zu sichern. Möglich wären tiefere Fundamente, erhöhte Plattformen oder der Rückzug des Gebäudes von der Kante.

Ob Cage diesen Weg gehen will, ist offen. Der Schauspieler hat in den vergangenen Jahren wieder verstärkt schauspielerisch gearbeitet und mit Filmen wie „Pig“ und „The Unbearable Weight of Massive Talent“ Kritikerlob erhalten. Seine Fangemeinde verfolgt daher nicht nur seine Filmprojekte, sondern auch die Entwicklungen in Malibu. Für die Nachbarschaft ist der Fall indes kein Einzelfall. Viele Anwohner fürchten, dass weitere Stürme ähnliche Schäden verursachen könnten.

Die Behörden haben nach den jüngsten Unwettern damit begonnen, gefährdete Zonen zu kartieren. Statiker sollen klären, welche Gebäude noch bewohnbar sind und welche evakuiert werden müssen. Für Nicolas Cage bedeutet das eine Phase der Ungewissheit. Weder der genaue Schadensumfang noch die endgültige Entscheidung über Abriss oder Sanierung sind bestätigt. Klar ist nur, dass die Naturgewalten an der kalifornischen Küste erneut gezeigt haben, wie verletzlich selbst Luxusimmobilien sein können.

Während die Aufräumarbeiten entlang des Pacific Coast Highway weiterlaufen, bleibt Cages Auffahrt ein sichtbares Zeichen der Zerstörung. Der Krater, das eindringende Wasser und die freigelegten Stützpfeiler dokumentieren, wie schnell ein jahrzehntelang aufgebautes Anwesen in Gefahr geraten kann. Die nächsten Prüfungen werden zeigen, ob das Haus eine Zukunft an diesem Ort hat oder ob die Küste es zurückerobert.

Source: MSN News

Le Grand Prix des Pays-Bas 2025 restera dans les mémoires comme une course folle, riche en rebondissements, en stratégies audacieuses et en incidents spectaculaires. Sur le circuit de Zandvoort, Lando Norris a signé une deuxième victoire consécutive au volant de sa McLaren, profitant d’un concours de circonstances favorable mais aussi d’un rythme impressionnant en fin de course. Derrière lui, Kimi Antonelli et George Russell complètent le podium, tandis que Lewis Hamilton a longtemps cru pouvoir arracher une place parmi les trois premiers.

Le départ a été le théâtre d’un accident majeur. Max Verstappen, quadruple champion du monde et héros local, a perdu le contrôle de sa Red Bull et a provoqué un drapeau rouge après un énorme crash. Heureusement, le Néerlandais est sorti indemne de sa monoplace, mais l’incident a coupé court à ses espoirs de victoire devant son public. La tension était à son comble dans les tribunes orange, où des milliers de supporters avaient fait le déplacement pour soutenir leur idole.

Un premier départ chaotique et un drapeau rouge

Dès les premiers mètres, la course a basculé dans l’inattendu. Verstappen, parti en première ligne, a soudainement perdu l’adhérence de sa Red Bull, probablement piégé par une piste encore froide ou par une erreur de pilotage dans le feu de l’action. Sa monoplace a violemment percuté les protections, entraînant l’interruption immédiate de l’épreuve. Le drapeau rouge a été déployé, et tous les pilotes ont regagné la voie des stands. Cet accident a rappelé à quel point Zandvoort est un circuit exigeant, avec ses virages relevés et ses courbes rapides qui ne pardonnent aucune approximation.

Avant le redépart, un autre incident a privé le peloton d’un concurrent : Oliver Bearman a abandonné durant le tour de formation. Le jeune pilote Haas n’a donc pas pu prendre part à la course, ajoutant une nouvelle désillusion à son début de saison. Au moment du départ effectif, Kimi Antonelli a réalisé une excellente mise en action et s’est emparé de la tête, tandis qu’Oscar Piastri parvenait à dépasser George Russell pour le gain de la troisième place. Lando Norris, de son côté, se maintenait dans le sillage des hommes de tête.

Antonelli solide leader, stratégies décalées

La course a ensuite pris une tournure stratégique. Le premier des favoris à rentrer aux stands fut George Russell, qui chaussa des pneus durs au 18e tour. Un tour plus tard, Oscar Piastri opta pour des médiums, tentant de couvrir son coéquipier. Au 22e tour, Kimi Antonelli, Lando Norris et Charles Leclerc rentrèrent à leur tour. L’Italien réussit à ressortir devant le champion du monde, conservant ainsi la tête de la course. Cette séquence illustrait parfaitement la confiance de Mercedes en son jeune pilote, capable de résister à la pression des McLaren et des Ferrari.

Antonelli, qui dispute seulement sa première saison complète en Formule 1, impressionne par sa maturité. À Zandvoort, il a géré ses pneus avec une précision remarquable, tout en maintenant un rythme élevé. Son duel à distance avec Norris a tenu les spectateurs en haleine. Derrière, Charles Leclerc et Lewis Hamilton tentaient de suivre le rythme, mais la Ferrari semblait légèrement en retrait sur ce circuit. L’écurie italienne misait sur une stratégie décalée pour revenir dans la lutte.

La deuxième vague d’arrêts et le retour de Norris

Au 42e tour, Antonelli effectua une nouvelle halte pour remettre des pneus durs. Norris fit de même six tours plus tard, mais ressortit derrière l’Italien. Le champion du monde en titre ne se découragea pas : il haussa son rythme et parvint à dépasser successivement le pilote Mercedes et Lewis Hamilton pour reprendre les commandes. Ce dépassement, effectué avec autorité, a démontré toute la détermination de Norris, qui ne voulait pas laisser filer une opportunité de victoire.

Peu après, la Virtual Safety Car fit son apparition suite à l’abandon d’Esteban Ocon. Les deux pilotes Ferrari en profitèrent pour rentrer et chausser des tendres, tout comme Antonelli et Piastri. Cette neutralisation a redistribué les cartes à une vingtaine de tours de l’arrivée. Les stratèges ont dû recalculer leurs plans, et les écarts se sont resserrés. La fin de course promettait d’être intense.

Consignes d’équipe et nouvelle Virtual Safety Car

Au 66e des 72 tours, George Russell laissa passer Kimi Antonelli suite à une consigne d’équipe. Le Britannique perdit donc sa deuxième place, mais Mercedes cherchait à maximiser les points de son jeune leader au championnat. Alors que Lewis Hamilton se rapprochait dangereusement de Russell pour aller chercher un podium, la Virtual Safety Car refit son apparition après une touchette entre les Williams d’Alexander Albon et Carlos Sainz. Cet incident mit fin aux espoirs de dépassement de Hamilton, qui dut se contenter de la quatrième place.

Devant, Lando Norris ne trembla pas. Le pilote McLaren géra parfaitement la pression et franchit la ligne d’arrivée en vainqueur. Il signe ainsi une deuxième victoire de rang, confirmant son statut de prétendant sérieux au titre mondial. Kimi Antonelli et George Russell complètent le podium, tandis que Lewis Hamilton échoue au pied du podium. Charles Leclerc, Oscar Piastri et les autres pilotes se sont battus pour les points restants.

Fernando Alonso et Aston Martin revigorée

À noter également la belle neuvième place de Fernando Alonso au volant d’une Aston Martin revigorée. Le double champion du monde espagnol a livré une course solide, marquée par une gestion impeccable de ses pneus et quelques dépassements bien sentis. Son équipe semble avoir trouvé des solutions pour améliorer la monoplace après un début de saison difficile. Cette performance encourageante pourrait annoncer une seconde partie de saison plus compétitive pour Aston Martin.

Classement du championnat du monde

Au championnat du monde, Kimi Antonelli compte désormais 59 points d’avance sur son coéquipier et Lewis Hamilton. Cette avance considérable place l’Italien en position de force avant la tournée européenne finale. George Russell, son coéquipier chez Mercedes, et Lewis Hamilton, chez Ferrari, devront trouver des solutions pour combler l’écart. Lando Norris, grâce à sa victoire, se replace également dans la course au titre, même s’il reste encore beaucoup de points à distribuer.

La saison 2025 est décidément pleine de surprises. Le crash de Verstappen à Zandvoort a rappelé que même le quadruple champion du monde n’est pas à l’abri d’une erreur. Red Bull, qui comptait sur son pilote vedette pour briller à domicile, repart avec zéro point et de nombreuses questions. L’écurie autrichienne devra analyser les causes de l’accident et réagir rapidement pour ne pas laisser Mercedes et McLaren prendre le large.

Zandvoort, un circuit unique et exigeant

Le circuit de Zandvoort est réputé pour ses virages relevés, notamment le célèbre virage 3, incliné à 18 degrés, qui met les pilotes à rude épreuve. Avec ses 4,259 kilomètres et ses 14 virages, il ne laisse aucun répit. Les dépassements y sont difficiles, ce qui rend les stratégies et les qualifications particulièrement cruciales. La ferveur des supporters néerlandais, vêtus d’orange, crée une atmosphère électrique unique en Formule 1. Cette année, l’absence de Verstappen dans la lutte pour la victoire a ajouté une dimension étrange à la course, mais le spectacle a été au rendez-vous.

Lando Norris a su tirer profit des circonstances pour s’imposer. Sa McLaren, équipée de pneus tendres en fin de course, semblait imbattable. Le pilote britannique a également bénéficié d’une excellente gestion de la pression, notamment lors des relances. Son coéquipier Oscar Piastri, bien que moins en vue, a marqué des points précieux. McLaren confirme ainsi sa position de force dans la lutte pour le championnat des constructeurs.

Les faits marquants

• Max Verstappen perd le contrôle de sa Red Bull au départ et provoque un drapeau rouge ; il sort indemne.
• Oliver Bearman abandonne durant le tour de formation.
• Kimi Antonelli prend la tête avant de la céder à Lando Norris.
• George Russell laisse passer Antonelli sur consigne d’équipe.
• Esteban Ocon abandonne, déclenchant une Virtual Safety Car.
• Alexander Albon et Carlos Sainz se touchent, nouvelle Virtual Safety Car.
• Lando Norris remporte la course devant Antonelli et Russell.
• Fernando Alonso termine neuvième avec Aston Martin.
• Antonelli possède 59 points d’avance sur Russell et Hamilton au championnat.

Rendez-vous dans deux semaines pour le Grand Prix d’Italie à Monza.

Source: RTBF News

• Course : Grand Prix d'Italie disputé le 6 septembre 2026 sur le circuit de Monza.
• Pilote : Charles Leclerc, au volant d'une Ferrari, a abandonné dès le deuxième tour après une sortie de piste à la Parabolique.
• Incident : après avoir défendu sa position face à Oscar Piastri, le Monégasque est sorti trop large et a percuté les protections en pneus. La course a été interrompue par un drapeau rouge.
• Santé : Leclerc n'a pas été blessé, mais a signalé des problèmes de vision temporaires à l'œil droit. Il a été autorisé à quitter le centre médical, avec des examens complémentaires prévus avant le Grand Prix de Madrid.
• Famille : son épouse Alexandra Saint Mleux est apparue dans les paddocks, arborant une robe longue rouge satinée et un ventre très arrondi.
• Grossesse : le couple, marié le 28 février 2026, attend son premier enfant, dont la naissance est envisagée pour 2027.
• Rappel : Leclerc s'était déjà crashé dans ce même virage lors de l'édition 2020, année du succès de Pierre Gasly à Monza.

Un dimanche contrasté entre joie familiale et accident sur la piste

Le Grand Prix d'Italie 2026 restera comme l'un des week-ends les plus riches en émotions pour Charles Leclerc. Sur le circuit de Monza, le pilote Ferrari devait composer avec une double actualité : la présence remarquée de son épouse Alexandra, enceinte, et un accident spectaculaire dès le deuxième tour. La journée avait pourtant bien commencé pour le Monégasque, qui pouvait compter sur le soutien de sa femme dans les paddocks. Mais la piste a rapidement rappelé à quel point la Formule 1 peut basculer en quelques secondes. Entre l'annonce d'un heureux événement et une sortie de piste impressionnante, le contraste n'a pas manqué de frapper les observateurs.

Alexandra Leclerc, future maman, illumine les paddocks de Monza

Avant le départ, toute l'attention n'était pas seulement tournée vers les monoplaces. Alexandra Saint Mleux, épouse de Charles Leclerc depuis le 28 février 2026, a fait une apparition remarquée dans les paddocks du Grand Prix d'Italie. La jeune femme portait une robe longue rouge satinée, qui mettait en valeur son ventre très arrondi tout en laissant ses épaules dénudées. Son sourire et son attitude rayonnante ont immédiatement attiré les regards. Pour les tifosi et les médias présents, il était impossible de manquer cette silhouette glamour, qui confirmait une nouvelle déjà annoncée par le couple : un bébé verra le jour en 2027, un an après leur mariage. Sur ses réseaux sociaux, Alexandra Leclerc partage régulièrement des clichés de sa grossesse, dévoilant une allure de sirène moderne, à la fois élégante et épanouie. Cette apparition à Monza a ajouté une touche personnelle et chaleureuse à un week-end autrement marqué par la tension sportive.

Le crash du deuxième tour : la Parabolique frappe encore

Sur la piste, l'ambiance a brutalement changé. Dès le deuxième tour, Charles Leclerc a été impliqué dans un accident. Selon les informations disponibles, le pilote Ferrari repoussait une attaque d'Oscar Piastri, alors que les deux hommes abordaient la Parabolique, l'un des virages les plus rapides et les plus exigeants du circuit italien. En défendant sa position, Leclerc est sorti trop large dans cette courbe légendaire. Il a perdu la maîtrise de sa monoplace, qui a terminé sa course contre les protections en pneus. Le choc a été violent, mais le Monégasque a pu s'extraire par ses propres moyens de son cockpit. Il est resté assis de longues minutes, entouré des équipes médicales, avant d'être conduit au centre médical du circuit. Cet incident a provoqué un drapeau rouge, obligeant les organisateurs à interrompre la course pour effectuer de lourdes réparations sur la piste avant un nouveau départ.

Des troubles visuels temporaires mais une inquiétude maîtrisée

Après l'accident, Charles Leclerc a été interrogé par la presse. Il a expliqué avoir rencontré momentanément des problèmes de vision au niveau de l'œil droit. Le phénomène s'est depuis résorbé, et le pilote a été autorisé à quitter le centre médical du circuit. Par précaution, des examens complémentaires sont toutefois prévus avant le prochain Grand Prix à Madrid. Cette décision montre que, même si le pilote n'a pas été blessé, le choc et les troubles visuels ont été pris très au sérieux par le staff médical. Pour Ferrari, l'essentiel était de voir son pilote sortir indemne d'un accident qui aurait pu avoir des conséquences plus graves. Pour Leclerc, futur papa, ce dimanche stressant s'est donc terminé avec plus de peur que de mal, mais avec la nécessité de surveiller de près son état de santé avant de reprendre le volant.

La Parabolique, ce virage déjà marqué dans l'histoire de Leclerc

Il faut dire que la Parabolique n'est pas un virage anodin dans la carrière de Charles Leclerc. En 2020, le pilote Ferrari s'était déjà crashé dans cette même courbe lors du Grand Prix d'Italie. Cette année-là, Pierre Gasly avait créé la sensation en remportant la course à Monza. Le souvenir de cet accident était encore présent dans les mémoires, et le nouvel épisode de 2026 n'a fait que renforcer la réputation de ce virage comme l'un des plus piégeux du calendrier. La Parabolique est abordée à très haute vitesse, et la moindre erreur de trajectoire peut se transformer en sortie de piste. Les pilotes doivent y trouver un équilibre parfait entre freinage, appui aérodynamique et maîtrise de la monoplace. Pour Leclerc, ce virage semble décidément associé à des moments forts, parfois heureux, parfois douloureux.

Monza, Ferrari et la pression des tifosi

Monza est un circuit à part pour Ferrari. Surnommé le temple de la vitesse, il accueille chaque année des milliers de tifosi venus soutenir la Scuderia. La pression y est immense, et chaque erreur est scrutée. Pour Charles Leclerc, courir à Monza avec une Ferrari représente à la fois un honneur et un défi. Le pilote monégasque y a déjà connu des moments de gloire, notamment une victoire en 2019, mais aussi des désillusions. En 2026, l'attente était encore plus forte, car le duo Leclerc-Ferrari cherchait à briller devant son public. L'accident du deuxième tour a donc été un coup dur, non seulement pour le pilote, mais aussi pour toute l'équipe. Le drapeau rouge et les réparations ont interrompu l'élan de la course, laissant les spectateurs dans l'attente d'un nouveau départ. Cet épisode rappelle que, sur ce circuit, la frontière entre l'exploit et l'accident est parfois infime.

Une carrière déjà riche en poles, victoires et rebondissements

Charles Leclerc n'est pas un novice en Formule 1. Arrivé dans la catégorie reine après un parcours impressionnant en GP3 et en Formule 2, il a rapidement été considéré comme l'un des grands espoirs de sa génération. Sa vitesse en qualifications, sa capacité à signer des poles et sa combativité en course lui ont permis de s'imposer comme un pilote de premier plan. Chez Ferrari, il a vécu des moments forts, avec des victoires prestigieuses et des batailles mémorables. Mais il a aussi connu des accidents, des abandons et des périodes plus difficiles. Cette capacité à rebondir après les épreuves fait partie de son caractère. Le crash de Monza en 2026 s'inscrit dans cette histoire faite de hauts et de bas, où chaque week-end peut apporter son lot de joies et de frustrations.

Mariage, grossesse et projets familiaux : une nouvelle dynamique

Sur le plan personnel, Charles Leclerc vit une période particulièrement heureuse. Marié à Alexandra Saint Mleux depuis le 28 février 2026, il se prépare à devenir père. Le couple attend un bébé dont la naissance est prévue en 2027. Dans un entretien récent, le pilote monégasque avait évoqué l'évolution de sa vie privée et les grands événements à venir hors des circuits. Il avait confié : « Ma vie va probablement pas mal changer d'ici 2027 sur le plan familial. Donc oui, on a tous vraiment hâte. C'est une période très excitante. » Ces propos pre

Source: Purepeople News

Elena Rybakina will become the world No. 1 for the first time in her career on Monday, following a hard-fought quarterfinal victory over Qinwen Zheng at the ongoing tournament. The 27-year-old Kazakh won 3-6, 6-1, 6-4, showing the blend of firepower and composure that has defined her rise. She replaces Aryna Sabalenka at the top of the rankings, marking a shift in the WTA hierarchy. Rybakina’s reaction to the milestone was characteristically understated: a shy smile, no fist pumps, no dramatic celebration. For a player who lets her racket do the talking, that silence speaks volumes.

At 1.84m, Rybakina possesses one of the most feared serves in women’s tennis. But her game is not just about size. It is about precision, aggression, and an almost stoic calm that allows her to take risks without spiraling. Those who have watched her closely describe her as shy, reserved, and quietly competitive. The contrast with Sabalenka, a fiery and expressive competitor, is stark. Yet Rybakina’s reserved demeanor is not a weakness; it is a tactical asset that keeps her opponents guessing and keeps her own emotions from interfering with her high-risk style.

An Exceptional Serve

The serve is Rybakina’s No. 1 weapon, and the numbers back it up. On the WTA website, not including the current tournament, she leads the top 10 in several key serving categories. She has struck 357 aces since the start of the season, more than any other player in the top 10. She has won 73.4% of points behind her first serve, tied with Linda Noskova for the best mark among the elite. And she has saved 65.6% of break points, also the highest in the top 10. Those statistics translate into a dominance on service games that has no equal on tour.

Former players and analysts have marveled at the weapon. One former top-10 player noted that Rybakina’s serve produces so many free points that it changes the entire complexion of a match. Even her second serve is excellent, which puts immense pressure on opponents when they serve. If they get broken, they know that breaking back is a monumental task because Rybakina is so reliable on her own delivery. Another analyst pointed out that the true brilliance lies not just in the aces but in her ability to place the ball almost anywhere in the box while maintaining power and consistency. That combination—placement, power, and consistency—is the winning trifecta.

The serve also sets up the rest of her game. Because she can win so many points outright, she is free to swing aggressively on return games and take chances. Opponents, meanwhile, feel compelled to over-press on their own serves, which leads to more errors and more break points for Rybakina. It is a psychological vise that she applies relentlessly.

Constant Aggression and Flat Power

Behind that fearsome serve, Rybakina applies constant pressure by attacking on virtually every shot. She refuses to drift far behind the baseline, instead looking to step in and take the ball early. Her shots travel fast through the air, very flat, so they do not bounce high and instead skid through the court. That style is particularly effective on faster surfaces, which explains why her first Grand Slam title came at Wimbledon. The grass rewards her flat, penetrating groundstrokes and her ability to rush opponents.

Analysts emphasize that her ball speed is exceptional. It is not the same heavy, spin-laden ball that some other top players produce; it is a flatter, skidding trajectory that gives opponents less time to react. Her technique allows her to generate phenomenal arm speed without appearing to force the effort. At contact, she looks smooth and unforced, unlike many power hitters who seem to muscle the ball. That efficiency is tied to her timing and her movement. Despite her height, she anticipates well and moves intelligently around the court. People sometimes fail to credit her footwork, but it is a crucial part of her ability to be aggressive without losing balance.

The risk of such a game is obvious: errors. Flat, aggressive hitters can rack up unforced errors when their timing is slightly off. But Rybakina has learned to manage that risk. She has worked on her targets, playing the lines less than she once did and giving herself a little more margin. That adjustment has allowed her to be more consistent without sacrificing her attacking identity. She still takes the ball early and looks to dictate, but she no longer goes for impossible winners on every shot. The result is a player who can sustain her aggression over three sets and through the later rounds of major tournaments.

Unshakable Calm

Perhaps the most distinctive feature of Rybakina’s game is her unshakable calm. On court, she is like a flat lake. There are no screams, no visible signs of joy or distress, no celebrations after winning a big point. Even after the quarterfinal win that sealed her rise to No. 1, she offered only a shy smile. Inside, there is no doubt that a lot is going on, but from the outside she does not flinch from the first point to the last.

That emotional stability is a perfect fit for her high-risk style. Because she plays with such small margins, she needs to remain even-keeled whether she is hitting winners or making errors. If she were prone to emotional highs and lows, her game would be far more volatile. Instead, she continues with the same intentions, the same aggression, the same belief. Opponents find this unsettling. Playing against someone who shows nothing—no frustration, no excitement, no fatigue—is annoying and disorienting. It gives them no emotional feedback to latch onto, no sense that they are close to a breakthrough. For Rybakina, it is a weapon as potent as her serve.

Analysts note that her calm also helps her physically. By not wasting energy on emotional outbursts, she stays fresh in long matches. She does not get caught up in the drama of a tight set or a bad line call. She simply moves to the next point. That professionalism and focus have allowed her to compete consistently at the highest level, even when her game is not at its absolute best.

What the Numbers Say

The statistical profile of Rybakina’s season is remarkable. Her 357 aces lead the top 10. Her 73.4% first-serve points won is tied for the best. Her 65.6% break points saved is the best. Those three categories—aces, first-serve points, and break points saved—are the pillars of service dominance. When a player leads all three, she is almost impossible to break. That is why Rybakina has been able to win matches even when her return game is not firing. She can simply hold serve and wait for one opportunity.

Beyond the serve, her aggressive return game and flat groundstrokes create additional pressure. She may not lead the tour in return games won, but she takes enough chances to keep opponents uncomfortable. The combination of a nearly unbreakable serve and a high-risk, high-reward return game makes her a nightmare matchup. Opponents know they may get only a few break points, and they must convert them because Rybakina will not give many away.

Historical Context and Career Highlights

Rybakina’s journey to No. 1 is the culmination of years of steady improvement. Born in Moscow, she began representing Kazakhstan in 2018, a move that gave her a new national team and a new platform. She broke through in 2022 by winning Wimbledon, becoming the first Kazakh player to win a Grand Slam singles title. That victory announced her as a major force, but it also came with questions about consistency and whether she could handle the pressure of being a champion.

She answered those questions in the seasons that followed. She won multiple WTA 1000 titles, reached the final of the Australian Open, and established herself as a perennial top-10 player. Her Australian Open title—the one referenced as her second major—cemented her status as a complete player, not just a grass-court specialist. She proved she could win on hard courts, on clay, and on grass, adapting her flat, aggressive game to different conditions.

Now, at 27, she is set to become world No. 1. The timing feels right. She has the experience, the weapons, and the temperament. She has learned to manage her risk, to trust her serve, and to remain calm in the biggest moments. The WTA tour has been searching for a consistent No. 1, and Rybakina’s blend of power and poise makes her a worthy occupant of the throne.

Impact on the Tour

Rybakina’s rise also signals a stylistic shift. She is not a baseline grinder who relies on endless rally tolerance. She is an attacker, a first-strike player who uses her serve and flat groundstrokes to end points quickly. Her success could encourage more players to develop aggressive, flat-hitting games. It also puts pressure on the rest of the tour to solve the riddle of her serve. Coaches will study her placement patterns, her second serve, and her break-point saves. But solving it is easier said than done.

Off the court, Rybakina’s reserved personality makes her a different kind of star. She is not a showman. She does not seek attention. She lets her tennis speak. That authenticity has won her respect from fans and peers alike. In an era of social media and constant content, her quiet confidence is refreshing. It also makes her a fascinating case study in how introverted athletes can thrive in individual sports.

As she prepares to take the No. 1 ranking, Rybakina will face new pressures. Every opponent will raise their level. Every match will be scrutinized. But if her career so far is any indication, she will handle it with the same ice-cold calm that has carried her to the top. The shy smile may be the only clue that she is enjoying the moment, but the scoreboard will tell the rest of the story.

Source: MSN News

A hybrid vehicle pairs an internal combustion engine with one or more electric motors and a small battery pack. The motor can start the car, provide extra power during acceleration, and capture energy normally lost when braking. That captured energy is stored in the battery and used later. In a conventional hybrid, the battery is not plugged into the electric grid; the engine and regenerative braking keep it charged. This is why hybrids are often called self-charging. The result is lower fuel consumption and reduced carbon dioxide emissions compared with a gasoline-only car, especially in stop-start city driving.

A battery electric vehicle removes the gasoline engine entirely. It uses a much larger battery to store electricity and power one or more electric motors. Because there is no combustion, the vehicle produces no tailpipe emissions, and its energy is usually expressed in kilowatt-hours. The driving experience is generally quiet and responsive, with strong acceleration from low speed. The biggest adjustment compared with a conventional car is refueling. An EV needs to be plugged into a charging point at home, at work, or on a public network, and charging takes longer than a typical gasoline fill.

Main types of electrified vehicles

• Mild hybrid: A small electric motor assists the gasoline engine but cannot move the car on its own. It improves efficiency and smooths stop-start driving.
• Full hybrid: The electric motor can drive the vehicle at low speeds, while the gasoline engine takes over for higher speeds and heavy acceleration.
• Plug-in hybrid: A larger battery is charged from a power outlet and offers a useful electric-only range, often covering short commutes.
• Battery electric: The vehicle runs entirely on battery power and must be recharged from the grid.

Key facts at a glance

• Hybrids and electric vehicles can lower fuel costs compared with many gasoline-only vehicles.
• Electric motors deliver torque quickly, which makes city driving and overtaking easier.
• Most conventional hybrids cannot be plugged in; they generate electricity through the engine and regenerative braking.
• Plug-in hybrids need regular charging to deliver their best economy.
• EV range has grown steadily, while public fast-charging networks continue to expand.
• Government policies are encouraging automakers to launch more electrified models across all segments.

How electrified vehicles reached the mainstream

The idea of electric propulsion is older than the gasoline car. In the late 1800s, electric cabs and private carriages operated in several cities. They were quiet and did not need cranking, but heavy batteries and limited range made them difficult as a mass-market product. The expansion of roads and cheap gasoline pushed automakers toward internal combustion, and by the early twentieth century EVs were mostly forgotten. The oil crises of the 1970s sparked new interest in efficiency, and later battery research created a platform for modern hybrids and EVs.

In the late 1990s, Toyota introduced the Prius, which became the best-known hybrid in the world. It demonstrated that electrification could reduce fuel use without asking drivers to change their habits. In the 2010s, Tesla and other newer manufacturers pushed battery range and performance forward, changing public expectations about electric cars. That shift encouraged large automakers to invest billions in electrified platforms, batteries, and charging technology.

Costs, charging, and long-term ownership

The list price of an EV has historically been higher than that of an equivalent gasoline car. That difference has narrowed as battery production scales and new factories come online. Government incentives, tax credits, and lower running costs can change the financial comparison. Electricity is generally cheaper per mile than gasoline, particularly when charging at home overnight on a residential tariff. Public rapid charging is less costly than petrol in many markets but more expensive than home charging.

Hybrids sit between the two worlds. A conventional hybrid usually costs more than a gasoline model but less than an EV, and it avoids the need for a charging routine. A plug-in hybrid can offer the best of both if the owner charges regularly. If the battery is left empty and the owner relies mainly on gasoline, a plug-in hybrid may return fuel consumption not much better than a regular car because of the extra weight of the battery.

Maintenance is another difference. EVs do not require engine oil changes because there is no internal combustion engine. Regenerative braking also means the brake pads are used less. Hybrids still need engine maintenance, but the engine may work less hard. Over time, owners can expect lower servicing costs for EVs, although battery wear remains an important factor. Most manufacturers offer long battery warranties, and modern battery management systems are designed to preserve capacity.

Emissions and wider environmental considerations

Electric vehicles produce zero tailpipe emissions, which improves local air quality in streets and cities. But the full picture depends on how the electricity is generated. A car charged on a coal-heavy grid causes higher upstream emissions than one charged using solar or wind power. Even so, studies show that in most countries a battery EV produces lower lifecycle emissions than a comparable gasoline car, especially as the grid becomes cleaner.

Hybrids can reduce emissions because they burn less gasoline, yet they still generate tailpipe emissions. For drivers who want a meaningful drop in emissions but cannot charge at home, a conventional hybrid is a practical step. For drivers with reliable charging, a plug-in hybrid can cover daily trips without using gasoline, while retaining a backup for long journeys. For those with home charging and access to public fast charging, an EV has the lowest running emissions.

Market trends that are reshaping the industry

The shift is not limited to passenger cars. Delivery vans, buses, taxis, and commercial vehicles are being electrified in many major cities. Automakers are redesigning their lineups around dedicated electric architectures rather than adapting old gasoline platforms. Battery chemistry is also evolving. Newer cells offer more usable range, faster charging, and better performance in cold weather. Charging networks are expanding along highways and in city centers, reducing the anxiety that once surrounded long-distance EV travel.

Policies in many countries now set dates for ending the sale of new gasoline and diesel cars. These dates affect mainstream brands and force the industry to move more quickly. But the pace of change is not uniform. In some regions, cost, infrastructure, and access to battery minerals remain serious obstacles. That is why hybrids and plug-ins remain important as an intermediate option, especially for motorists with no private parking or limited public charging near their homes.

Source: TechRadar News

President Trump responded forcefully to the findings in messages on his Truth Social platform. He accused residents and local officials who block data centers of “choosing poverty” rather than accepting economic benefits from server farms. He added that such facilities “throw off tremendous amounts of money” and maintained that turning them away would damage the country’s competitive standing in the global artificial intelligence race.

Local concerns are widespread

The poll’s finding of 69 percent opposition is one of the strongest signals of “not in my backyard” sentiment associated with critical infrastructure in recent years. It reflects more than a general dislike of large buildings; it captures a sense among many residents that data center development is planned for their counties without enough transparency or accountability.

Unlike a factory or warehouse, a modern AI data center can run continuously for decades. It is full of powerful processors that generate intense heat. The facility must therefore be cooled around the clock, often using air handlers that produce low-frequency sound or water that evaporates in cooling towers. Power demands can be equally impressive. A single hyperscale site can draw as much electricity as a medium-sized town. The combined resources needed to train advanced AI models have escalated so rapidly that some grid operators compare the projected load from data centers to a new state entering the grid.

That heavy resource consumption has created unlikely alliances. Farmers worried about water scarcity have joined suburban residents concerned about transformer noise and environmental groups have partnered with rural conservatives who object to corporate subsidies. School boards have raised questions about whether new tax revenue will cover the strain on roads and emergency services. All these factors appear to have pushed overall opposition in the new poll past two-thirds.

Trump’s message to resisting communities

Rather than soften his pitch, Trump chose to draw a hard line. His Truth Social posts are aimed at supporters who might otherwise side with local jurisdictions. He argued that America’s future and national security depend on the willingness of cities and counties to host the physical backbone of artificial intelligence.

He wrote that communities blocking server farms are choosing poverty over prosperity. The phrase “tremendous amounts of money” appeared in his description of tax revenue and business activity generated by the data center industry. Trump’s intervention is noteworthy because land-use decisions usually fall to state and local authorities. He placed the debate in an international context, framing local objections as a danger to American competitiveness against China and other nations.

Yet local voices counter that there is no contradiction between supporting AI and wanting responsible placement of massive facilities. They argue that national urgency should not override community consent for projects that will affect residents’ daily lives for a generation.

The AI building boom

Behind the political combat lies a technological rush. Generative products such as chat assistants, coding copilots, image generators and autonomous agents are consuming processor capacity at an almost limitless pace. Major technology firms have announced billions of dollars in capital spending on new cloud regions and data centers. Utilities have placed moratoriums on new industrial service connections in several areas because they cannot generate or transmit power fast enough to satisfy incoming requests.

State officials are also competing for investment. Indiana, Ohio, Georgia, South Carolina and Arizona have revised tax

Source: TechRadar News

Program officials describe the engine choice as a deliberate move to lower risk without lowering capability. The F-16 engine has one of the widest maintenance and logistics footprints in the Western fighter fleet. Thousands of examples have been built, repair depots know every major component, and operator experience is already global. By placing that engine into an AI-piloted vertical takeoff and landing airframe, designers can concentrate their development budget on autonomy, flight controls, and vertical-lift integration rather than creating an unproven propulsion system from scratch.

Key facts at a glance:

• The aircraft will use a proven 36-year-old engine from the family originally developed for F-16 fighters.
• It is described as the world's first AI-piloted VTOL fighter, combining vertical takeoff and landing capability with fixed-wing speed.
• The engine is an afterburning turbofan with a long combat record in air-superiority, strike, and suppression missions.
• The onboard AI pilot is expected to manage takeoff, landing, and combat maneuvers, reducing risk to human aviators.

A proven powerplant from the F-16 era

To understand why an older engine is making news, it helps to remember why the F-16 needed a revolutionary powerplant in the first place. The F-16 was designed to be small, cheap, and highly maneuverable, but early production models were powered by engines that sometimes struggled with the demands of aggressive dogfighting and rapid throttle movement. Military planners wanted an alternative that could deliver more thrust and better response across a wider range of flight conditions.

The engine now linked to the AI-piloted VTOL project emerged from that push for options. It belongs to the generation of afterburning turbofans that gave the F-16C and F-16D greater combat flexibility. By using a larger fan and a more efficient core, it produced more thrust at lower temperatures and with better fuel economy. That made the aircraft harder to stall in a fight, quicker to accelerate, and more capable with heavy weapon loads. The design also opened the door for the F-16 to carry out missions far beyond the close-range air-combat role originally imagined for it.

What makes the engine especially attractive today is not merely its performance but its maturity. In aviation, a new engine can require years of testing before it is accepted for combat. This engine is already past that barrier. It has been run, rebuilt, modified, and improved for more than three decades. Operators around the world maintain the skills needed to service it. The spare-parts pipeline is dense. In a military environment where reliability is often more important than raw speed, choosing a mature engine can be the safest possible decision.

Why an AI-piloted VTOL fighter changes the equation

The aircraft itself represents a more recent ambition: to remove the pilot from missions that are too repetitive, too dangerous, or too likely to be contested. A vertical takeoff and landing fighter does not need a long runway, a catapult, or an arresting gear. It can operate from small decks, clearings, roadways, or forward bases. That flexibility matters because modern anti-access systems threaten the large airfields on which traditional fighters depend.

But VTOL flight places brutal demands on any propulsion system. Hovering consumes enormous energy. Transitioning from a vertical hover to wingborne forward flight requires precise control of thrust, pitch, and attitude. If anything goes wrong during that transition, the margin to recover can be measured in seconds. An engine with a slow throttle response can make a hover unstable. An engine without enough reserve thrust can leave the aircraft unable to arrest a descent or climb away from an obstacle. The chosen F-16-derived powerplant offers the high thrust-to-weight ratio and rapid response that vertical flight requires.

There is also a historical connection between fighter engines and vertical flight. Some of the first attempts at vertical takeoff fighters used powerful jet engines to drive lift fans, direct exhaust downward, or rotate nozzles. Those efforts taught engineers that the engine is only part of an integrated flight-control problem. It must work with the airframe's flight computer, sensor suite, and control surfaces to keep the aircraft balanced near zero airspeed. That lesson is now being applied to a machine whose entire flight envelope may be managed by an artificial intelligence pilot.

What the AI pilot adds to a fighter airframe

The phrase AI-piloted can be misleading. Many remotely piloted aircraft already take off and land without direct human stick inputs. Autopilots can fly headings, altitudes, and navigation routes. The new development is a higher degree of decision-making autonomy. An AI pilot does more than maintain a steady course; it is being designed to interpret the battlefield, choose maneuver sequences, and execute tactical actions in real time.

For a vertical takeoff and landing fighter, that means the AI must handle the hardest parts of flight without relying on a ground controller to assist every movement. It must sense wind, temperature, and obstacles during takeoff. It must decide when the aircraft has enough airspeed to transition from vertical to horizontal flight. It must also know how to abort that transition if an engine malfunction, a control-surface failure, or an unexpected threat appears. These are not simple programming tasks. They require machine-learning systems trained on large amounts of flight data, as well as conventional software built to guarantee safe behavior.

Military autonomy also raises questions about trust. A pilot in a cockpit can feel the aircraft's vibrations, hear its engine, and instantly judge whether something is wrong. An AI pilot relies on sensors and electronic signals. For this reason, the first AI-piloted VTOL fighter is unlikely to be given total freedom from the beginning. Developers will almost certainly test the autonomy in progressively more difficult stages, beginning with ground taxi tests, then tethered hovers, then full transitions, and finally combat-like maneuvers.

A military role that does not need a runway

Defense analysts see several roles for an AI-piloted VTOL fighter. The most obvious is expeditionary operations from small ships. Many navies operate vessels with flight decks large enough for helicopters but not for conventional jet fighters. An AI-piloted fighter with vertical takeoff and landing capability could give those ships a fast, armed air asset without requiring a flattop carrier or a catapult system.

A second role is distributed operations in the Pacific and other island-heavy regions. Rather than concentrating fighter aircraft at a few large bases, a VTOL fighter could operate from temporary runways, austere airstrips, and even cleared sections of highway. That makes it harder for an adversary to target the aircraft before it can get airborne. The same engine that once defended NATO airspace may spend its final years operating from remote islands or frontline expeditionary sites.

The aircraft could also support special operations and search-and-rescue missions. Hovering capability allows it to loiter over a location, observe a target, and engage with precision weapons. Should the mission require retrieval of a downed pilot or a time-sensitive target, the AI can respond faster than a remotely piloted system waiting for instructions from a ground station. The lack of a human pilot also means the airframe can be designed around sensors and fuel rather than around cockpit ergonomics, ejection seats, and pilot life-support systems.

What a 36-year-old jet engine can still provide

It is tempting to judge an engine by its age, but jet engines do not age like consumer electronics. Military engine designs are continually modified through service-life extensions, new turbine materials, upgraded digital controls, and improved cooling technologies. An engine that first flew decades ago can still embody a sophisticated balance between airflow, compression, combustion, and thrust. The engine in question has also proven that it can absorb battle damage and still return the aircraft to base, a quality that autonomous systems cannot easily replicate because the aircraft has no pilot aboard.

The choice of a proven engine could also accelerate certification. In military aviation, a new engine-airframe combination requires extensive ground testing, flight testing, and environmental testing. By using an engine with a long safety and performance history, the VTOL fighter team can focus on the unknowns that truly matter: autonomy software, vertical flight control laws, and human-machine trust. Fewer unknowns usually mean a shorter path to deployment.

There are challenges, however. A 36-year-old engine design may not offer the same fuel efficiency as newer engines scheduled to enter service in the coming decade. It may require more maintenance hours per flight hour. Its exhaust temperature might limit the kinds of deck coatings or surfaces from which it can operate in vertical mode. And because the aircraft is unmanned, any engine failure that results in the loss of the airframe may still carry high financial consequences, especially if the aircraft is carrying advanced sensors or weapons.

There is also the broader question of whether the world needs an AI-piloted fighter today. Opponents of such programs argue that autonomous aircraft can be jammed, spoofed, or hacked. They note that no algorithm has yet demonstrated the full creativity and judgment of an experienced fighter pilot in a chaotic dogfight. Supporters counter that human pilots are becoming the most vulnerable element of a combat aircraft. They point to the growing cost of pilot training, the need to strike targets in high-risk environments, and the possibility of using swarms of cheaper autonomous aircraft to overwhelm defenses.

That debate will continue long after the first prototype reaches vertical hover. What is clear is that the project has chosen a deliberately conservative path in one of the most exciting, experimental corners of aviation. Instead of betting on a mysterious powerplant or a futuristic energy source, the designers are betting on a familiar engine that has already survived generations of pilots, wars, and technical revolutions. The next generation of that engine will now be asked to carry a different kind of fighter into a different kind of battlespace.

If the program succeeds, it will demonstrate that revolutionary progress in military aviation does not always require sending more money to an engine factory to build a new core from scratch. Sometimes the fast path to the future is found by taking a proven source of power from an older fighter and handing it to an artificial mind that has never known

Source: TechRadar News

Her trajectory is anything but accidental. Anitta grew up in Honório Gurgel, a suburb of Rio far from the country's wealthiest neighborhoods. She was raised by a single mother, began singing at age eight and took dance classes from the age of eleven. The favelas and working-class communities of Rio are not only her origins; they are also the cultural source she always claimed. She understood early that music is an art but also an industry and that she had to learn to master both dimensions if she wanted to escape invisibility.

From the Favela to the Global Mainstream

Anitta achieved national fame in 2012 with “Show das Poderosas”, a song that mixed funk, pop, dancehall, and a fierce feminine energy. Her subsequent works, with titles like “Veneno”, “Medicina”, “Downtown”, or “Loco”, have accumulated hundreds of millions of views. The music video for “Loco”, released at the beginning of 2021, passed over ten million views in less than two weeks. The singer displays in this clip a snow-covered landscape and a series of feminine bodies, including with trans woman La Demi, long before other Latin artists have made this kind of explicit commitment.

Anitta however remains first and foremost a musical phenomenon. She collaborated with Cardi B, Major Lazer, Snoop Dogg, and Becky G. With Becky G, she launched the video “Banana”, a colorful and eccentric work, named after the pop-art aesthetic associated with Andy Warhol. Through these collaborations, she has done what no Brazilian artist had done for years: making baile funk travel beyond the frontiers of Brazil. She said she wanted to export this music genre born in the favelas to show the world that it is art when it is done well.

Unlike many performers who pretend that the commercial dimension is secondary, Anitta has always assumed the financial aspect of her career. At the beginning of her career, she carried out market research to understand

Source: Terrafemina News

La démence est une maladie redoutable. Elle efface lentement les visages, les lieux et parfois même les moments les plus importants. La mère de cette jeune femme ne sait plus toujours pourquoi elle se trouve dans une pièce. Il lui arrive de demander où elle va, pourquoi elle a préparé ses bagages, ou ce qu’elle est en train de vivre. Pourtant, quand le nom de Jimin est prononcé, tout s’éclaire. Elle sait immédiatement de qui il s’agit, ce qu’il représente et pourquoi il compte tant pour elle.

Une annonce qui provoque une joie immense

Dans une vidéo partagée sur les réseaux sociaux, la fille annonce à sa mère : « Maman, j’ai acheté des billets pour qu’on puisse aller voir Jimin. » Ces quelques mots suffisent à illuminer le visage de la mère. Son sourire, son enthousiasme et son émotion montrent à quel point ce chanteur occupe une place particulière dans son cœur. Le simple fait d’évoquer cette perspective rend la maladie presque invisible pendant quelques secondes.

Ce genre de réaction n’a rien de surprenant pour les proches de personnes atteintes de démence. La maladie détruit certains types de mémoire, mais elle épargne parfois les souvenirs associés à des émotions fortes. La musique, en particulier, peut réactiver des zones du cerveau liées à la mémoire autobiographique et aux sentiments heureux. Jimin, avec sa voix douce et sa présence scénique, représente pour cette mère un univers de joie, à la fois personnel et collectif.

Le voyage vers Chicago : des hauts et des bas

Le voyage en avion a été marqué par ces allers-retours constants entre la lucidité et l’oubli. Pendant une partie du trajet, la mère répétait avec joie qu’elle allait voir Jimin. Elle fredonnait des airs et déclarait plusieurs fois : « Je vais voir mon Jimin. » Puis, quelques minutes plus tard, elle pouvait se tourner vers sa fille et demander pourquoi elles étaient parties, où elles allaient, ou ce qu’elles comptaient faire avec leurs valises. À ces instants de doute, la présence de sa fille était essentielle. Sans faire de grandes explications, il suffisait parfois de reparler du concert pour que la mère retrouve le fil de cette aventure.

Dans ces moments de confusion, la patience est clé. La fille a choisi de ne pas corriger ou brusquer sa mère. Elle l’a accompagnée, rassurée, et a recréé le contexte du voyage à chaque fois que cela était nécessaire. Ce comportement est souvent recommandé par les professionnels de santé qui accompagnent les familles confrontées à la démence. Laisser la personne évoluer dans sa temporalité, l’aider à se reconnecter aux émotions positives, et éviter les confrontations permet de préserver un lien affectif précieux.

Le concert de BTS et l’accès VIP

Arrivées à Chicago, la mère et la fille ont pu profiter d’un accès VIP pour le concert de BTS. Elles ont assisté à une représentation impressionnante, dans une salle remplie de fans venus du monde entier. Pour la mère, être dans cette foule et voir Jimin sur scène représentait sans doute beaucoup plus qu’un simple spectacle. La musique, la lumière, les milliers de voix qui chantaient en cœur et l’énergie du groupe formaient un cadre unique, propice aux émotions intenses.

BTS, formé par sept membres, est l’un des groupes les plus influents de la planète. Jimin, au côté de RM, Jin, Suga, J-Hope, V et Jungkook, séduit par sa voix expressive et ses talents de danseur. Le groupe s’est fait connaître au-delà des frontières coréennes grâce à des titres comme Dynamite, Butter ou Permission to Dance. Les concerts de BTS sont réputés pour leur scénographie soignée, leurs chorégraphies précises et l’interaction avec le public. Pour de nombreux fans, ces événements représentent un moment de communion inoubliable.

Le fait que la mère et la fille aient pu y assister ensemble donne une profondeur supplémentaire au voyage. C’est un souvenir commun que personne ne pourra leur enlever, même si la maladie risque d’en effacer une partie. Les photos prises sur place, les objets collectés et les moments partagés deviennent autant de balises émotionnelles.

La boutique éphémère et les souvenirs matériels

Autour du concert de Chicago, BTS avait installé une boutique temporaire destinée aux fans. La fille y a conduit sa mère afin qu’elles puissent toutes les deux explorer l’espace consacré au groupe. Elles y ont trouvé de nombreux produits dérivés à l’effigie des membres, notamment de Jimin. La jeune femme a choisi plusieurs objets pour décorer la chambre de sa mère. Elle espérait ainsi créer un environnement rassurant, rempli de références positives. Les photos prises lors de cette journée montrent la mère arborant un T-shirt au nom de Jimin, un grand sourire aux lèvres.

Ce n’est pas la première fois que des fans utilisent les produits dérivés ou les œuvres de BTS pour créer un lien avec un proche souffrant de troubles de la mémoire. L’objet visible, le vêtement, l’affiche ou encore la chanson peuvent servir de déclencheur à des conversations et à des moments de partage. La chambre décorée devient alors un espace de réconfort. La présence quotidienne du nom et du visage de Jimin permet peut-être à cette mère de conserver chaque jour un point d’ancrage.

Une gratitude profonde envers Jimin

Dans le message publié sur les réseaux sociaux, la fille s’adresse directement à Jimin. Elle écrit : « Je ne sais pas comment je pourrai un jour remercier Jimin pour ces petits moments que j’ai pu partager avec ma mère. » Cette phrase résume une gratitude immense, non pas envers une idole lointaine, mais envers un artiste dont la simple évocation maintient un lien vivant entre une mère et sa fille.

La dimension paradoxale de cette histoire est qu’elle ne repose pas sur une rencontre personnelle avec Jimin. Le pouvoir de l’artiste agit ici à distance, par le biais de la musique et de l’admiration. Pour la mère, Jimin est peut-être associé à des moments heureux, des chansons écoutées en famille, des souvenirs de sa fille. La présence de Jimin dans sa mémoire stable devient ainsi une présence de son enfant, de leur relation et de leur amour partagé.

Une communauté de fans unie autour de l’émotion

Partager cette histoire sur les réseaux sociaux a suscité de nombreuses réactions. Dans la sphère des amateurs de BTS, appelée ARMY, les témoignages de reconnaissance envers la musique du groupe se multiplient régulièrement. Beaucoup racontent comment les chansons de BTS les ont aidés à traverser des périodes difficiles. La communauté offre un espace de soutien où les expériences personnelles sont vécues comme des ressources collectives.

Cette histoire rappelle que la maladie d’Alzheimer et les troubles associés ne suppriment pas complètement la personnalité d’une personne. Les goûts musicaux, les attachements affectifs et les identifications culturelles peuvent continuer d’exister après l’installation des troubles cognitifs. Les familles qui découvrent cela vivent souvent des moments de grâce, quand un souvenir ressurgit grâce à une chanson, une photo ou un objet familier.

Dans le cas de cette mère et de sa fille, le fait d’avoir participé à une expérience aussi forte que le concert de BTS a créé une nouvelle strate de souvenirs. Même si certains seront oubliés, l’émotion ressentie ce jour-là restera sans doute inscrite quelque part dans la mémoire profonde de la mère. Comme la musique de la K-pop résonne dans les stades, elle résonne aussi dans les cœurs de ceux qui y cherchent du réconfort.

Quand la culture populaire croise la lutte contre l’oubli

Ce récit dépasse le simple cadre du divertissement. Il pose une question importante sur la manière dont la culture populaire peut entrer dans nos vies et y remplir une fonction que nous ne soupçonnons pas toujours. Les chansons préférées, les chanteurs admirés, les idoles aimées pendant l’adolescence ou l’âge adulte font partie de notre patrimoine affectif. Lorsqu’une grande partie de ce patrimoine devient inaccessible à cause d’une maladie, il peut subsister des îlots intacts auxquels se raccrocher.

Pour cette raison, les psychologues et les spécialistes de la démence s’intéressent de près à la musicothérapie et aux protocoles de réminiscence. La stimulation musicale est capable de provoquer des réactions émotionnelles et parfois même motrices chez des personnes qui ne répondent plus à d’autres types de sollicitations. Le fait que le nom Jimin soit prononcé avec certitude par une femme qui ne se souvient plus toujours de son quotidien en est une illustration éclatante.

Le rôle de cette jeune femme dans l’organisation du voyage, son accompagnement quotidien et sa manière de transformer une annonce d’achat des billets en un événement émotionnel spécial montrent aussi l’importance de la transmission familiale. Elle n’a pas simplement acheté des billets ; elle a offert à sa mère un moment suspendu, où la maladie perdait temporairement du terrain.

Un attachement qui dépasse la scène

Les images de cette mère en train de sourire devant l’objectif, les yeux brillants et le corps vêtu d’un T-shirt Jimin, constituent un témoignage touchant de la vie qui continue malgré tout. La démence, avec son cortège de pertes et de confusions, n’a pas réussi à volatiliser la passion de cette femme pour la musique de BTS. Au contraire, cette passion est devenue un outil précieux pour sa fille, une façon de la rejoindre là où elle se trouve encore.

Chaque voyage, chaque photo et chaque chanson partagée deviennent des actes de résistance face à l’effacement mémoriel. Dans l’avion, dans la salle de concert, dans la boutique éphémère, la mère et la fille ont accumulé des instants d’une intensité rare. Ils témoignent de ce que la relation entre un parent et son enfant peut traverser les épreuves les plus dures.

Quant à Jimin, il apparaît ici comme un élément déclencheur de bonheur. Sans doute n’imagine-t-il pas, sur scène, combien sa simple présence musicale peut illuminer la vie d’une personne atteinte de démence. Cette nouvelle preuve d’amour de la part des fans de BTS montre que le groupe a depuis longtemps dépassé le stade d’un simple phénomène musical.

Source: K-Sélection News

The open implementation described below connects a private Git-hosted Markdown vault to multiple LLM surfaces through a single remote Model Context Protocol, or MCP, endpoint. The system, vault-mcp, treats a folder of notes as a linked Obsidian vault and exposes it to claude.ai on the web, desktop apps, phones, and Claude Code in the terminal. The project also produced Zod AOT, an open-source ahead-of-time schema compiler, needed because edge platforms forbid runtime code generation, a restriction that silently disables the fast path of the TypeScript validation library Zod.

Key facts

• Coding AI assistants already have mature context tooling based on project files, repository maps, memory banks, and general-purpose memory frameworks.
• Personal context tools are largely locked to individual vendors, with no portable memory standard.
• The proposed fix separates context storage from assistant vendors, using user-owned Markdown stores connected through the open MCP protocol.
• vault-mcp implements the idea with a GitHub repository, an Obsidian vault, and a remote MCP server that works with multiple assistant surfaces.
• Edge platforms ban runtime code generation, which disables Zod's hidden just-in-time fast path; Zod AOT restores speed by compiling schemas at build time.
• As MCP spreads into e-commerce and new device categories, user-owned context can make assistants more personal, portable, and durable.

Coding context is almost solved

The modern developer has a rich set of context tools. Convention files are the simplest layer. Claude Code uses CLAUDE.md; cross-tool projects use AGENTS.md; Cursor and GitHub Copilot have their own rule files. These project-level files tell an agent how the codebase is organized and how to behave. Above them sit structural tools: repository maps compress the shape of a codebase into a context window, while code indexers and code-search integrations let an agent retrieve the exact file it needs. Persistence layers add memory across sessions. Cline's Memory Bank pattern stores structured notes, Cursor offers session memories, and Claude Code maintains a memory directory. Teams building agents on their own can choose general-purpose memory frameworks such as mem0, Letta, and Zep, which provide retrieval pipelines, ranking, and temporal knowledge graphs that track when a fact stopped being true.

The coding context ecosystem is crowded because the problem is well shaped. Code lives in repositories, is versioned, and has a structure that tools can exploit. The context is already stored independently of any single LLM prompt, and assistants can be pointed at it without losing provenance.

Personal context is an afterthought

Outside the terminal, the picture changes. Conversational use is the larger everyday LLM pattern for many people. Users ask questions grounded in their own situation: recent bloodwork trends, real portfolio exposure, previous vendor evaluations, or the promises made in earlier emails. Useful answers depend on personal context, yet the tooling is thin. Where memory exists, it usually lives inside ChatGPT, Claude, Gemini, or another assistant. Each tool remembers within its own boundary but not across them. Memory does not travel. Portability, when offered at all, arrives as a one-directional import controlled by the destination vendor. A person who switches assistants or uses several will see context fragment into inconsistent silos.

The extensible memory frameworks used by coding agents do not solve this either. mem0, Letta, and Zep are developer infrastructure for building agent products, not personal data stores. They keep memory in their own repositories, recreating lock-in one level down: a user can leave ChatGPT's silo but enter a startup's silo. The deeper question is why personal context should live inside any assistant at all. The answer proposed here is to decouple storage from the vendor.

A user-owned context layer

The basic architecture is plain and portable. Personal context is kept as plain text Markdown in a private GitHub repository, organized as an Obsidian vault, and exposed through an authenticated MCP endpoint. MCP matters because it is an open protocol rather than a proprietary SDK. A single server can serve claude.ai, the desktop app, a mobile app, and Claude Code in the terminal, allowing all of them to read and write the same notes. Any MCP-capable client can join in principle, so no vendor license is needed for interoperability. The neutrality of the design comes from the storage format, not from a platform. The corpus remains plain, open files in a Git repository. Leaving a vendor costs exactly git clone. A user never has to file an export request or wait for companies to agree on a migration format.

Transparency also improves. What an assistant knows about a user is a folder of readable, diffable files. Vendor memory products are now beginning to offer similar transparency, but user-owned storage had the property from the start. GitHub and Markdown are choices, not requirements. The principle asks only that context is outside the assistant in a store the user controls, accessible over MCP. A Google Docs store with a Docs MCP server works, as does a Notion workspace with a Notion MCP server. Each choice changes trade-offs: Git-hosted plain text maximizes portability and auditability, while a hosted workspace adds convenience. The separation between storage and assistant is the part that creates durable freedom.

Security by conservative defaults

Implementation choices carry more than plumbing because they define the security posture. vault-mcp runs on serverless edge infrastructure and uses the GitHub API as transport, so no personal computer has to remain powered on and the project fits into free tiers. Writes are append-only in a specific sense: an assistant can create or overwrite a note but cannot delete one. Every change becomes a Git commit, which makes the history auditable and enables rollback of any write. Authentication separates who can connect from what the server can touch, using two scoped credentials to limit damage if one leaks.

There is another subtle risk. When a personal knowledge base is exposed to LLMs, the notes themselves become untrusted inputs. A malicious instruction buried inside a note could act as a prompt injection the moment a model reads it. The server therefore treats retrieved notes as data rather than as instructions. Append-only behavior and path restrictions act as backstops. These are conservative defaults because the payload is small in volume but enormous in sensitivity: it is everything a person knows.

The edge constraint that forced a new tool

Serving the context vault from edge infrastructure was the right choice for zero maintenance, but it exposed an issue in a surprising layer: request validation. An MCP server is a public API. Every assistant tool call arrives as untrusted input and must be validated before touching storage. TypeScript developers usually reach for Zod, and the official MCP SDK is built around it. Zod's current performance comes from a hidden just-in-time compiler. The first validation of a given object shape generates specialized JavaScript for that shape and compiles it at runtime; later validations use the specialized code. Edge platforms such as Cloudflare Workers consider runtime code generation a security hazard in multi-tenant environments and forbid it. Zod detects the platform and falls back to a slower interpreted route.

That environment was exactly the one where the ecosystem's most common validator could not run at full speed. The solution followed Zod's own insight: schemas are static. They are written at development time and do not change during server execution. If specialized code cannot be generated at runtime, it should be generated at build time. The result is Zod AOT, an open-source compiler that transforms Zod schemas into flat JavaScript validation functions during the build. Because no code is produced during execution, the compiled validators work on edge runtimes and any other restricted environment. Because no compilation happens on the first request, there is no cold-start penalty. Benchmarks cited in the project show compiled validators checking complex nested objects up to sixty times faster than Zod's normal path, with a wider gap on edge platforms where Zod's fast path is unavailable. Schemas that embed arbitrary functions are partially compiled, and the remaining parts fall back to Zod unchanged. vault-mcp now validates every tool call with precompiled validators.

MCP turns chat into a way to act

The case for separating storage from vendor becomes stronger as MCP adoption spreads. MCP is becoming the standard method for connecting LLMs to external services. Every service that exposes an MCP endpoint makes chat a place where concrete work happens. E-commerce is one example. If a store exposes product search and checkout over MCP, an assistant can find a product, compare it with alternatives, and make a purchase in one conversation. A user-owned context layer makes the advice genuinely personal because the assistant can also read notes about sizes, budgets, past purchases, and disappointments. It can even recommend against buying, something a recommendation engine built for a seller is unlikely to do.

Service providers may not all welcome this shift. An MCP checkout shortens the time spent on a store's website and routes around the merchandising and advertising systems that generate e-commerce revenue. Some platforms will defend a destination site; others may decide it is better to be available where the customer already is. The user retains the context layer regardless of which path the market takes.

MCP also delivers device independence. A remote context vault does not assume a laptop or a smartphone. It is exposed through a protocol rather than an application-specific system. Future device categories such as smart glasses, wearables, home devices, and vehicle assistants can access the same vault in the same way as desktop clients today. There is nothing to migrate and no per-device memory that must be rebuilt from an empty state.

What this design accepts

The approach is not free. A user-owned vault does not offer ranked retrieval or a temporal model that knows when an individual fact has expired. Dedicated memory frameworks remain better for high-volume, automated recall. For one person's corpus of hundreds of notes, it is reasonable to let the assistant decide through tool use which notes to read, and that is the scale the system targets. Markdown-over-MCP servers and git-backed vault servers already exist. What is distinctive is the specific combination of user-owned storage, open-protocol access, no always-on host, conservative write semantics, and reuse of a build-time validator. The assistant is replaceable; the context is not.

Personal context affects finance, health, work, relationships, and decisions made over decades. Because edge platforms increasingly shape modern services, the work that went into Zod AOT offers a general solution for any team that validates data on the edge. The same separation principle that lets a lifelong knowledge store survive vendor changes helps embedded validation code run in environments where JIT-style shortcuts are illegal. For high-volume agent memory, a simpler vault should not replace dedicated infrastructure; for personal use, the trade is compelling. The last line of defense is not a clever ranking model but the knowledge that the files can be read and moved by their owner. That distinction keeps the proposal grounded in uncomfortable but manageable limitations.

Source: TNW | Contributed News

The valuation jump could be read as more of the same froth that has defined AI infrastructure financing over the past two years. But the more telling detail is that the multiple did not increase. In May, Cognition raised over $1bn at a $26bn valuation with reported run-rate revenue of $492m, implying a price of roughly 53 times revenue. The new $48bn valuation against almost $900m in run-rate revenue implies approximately the same multiple. Investors paid the same price per unit of revenue, but for a business that is roughly twice as large, at least as measured by the company's own figures.

The distinction matters because bubble-like dynamics are usually associated with multiple expansion. When investors pay more for each dollar of revenue even though growth is static, valuation theory begins to look detached. Here, the valuation re-rating tracked the reported business. That is not a sign of conservatism; 53 times revenue is still an extraordinary price, but it is a different risk profile from one in which the multiple itself is expanding.

Why the flat multiple is the real story

The same arithmetic could be described in simple terms: in May, investors valued Cognition at $26bn, and the company told them it had run-rate revenue of $492m. Now, the company says run-rate revenue has grown to almost $900m, and investors value the business at $48bn. If both prices are divided by their respective revenue claims, the implied multiple is roughly 53 times in each case.

That raises a question: did investors become more aggressive? On the available numbers, no. They simply paid the same multiple for a company with more reported revenue. That is not the profile of a market in which the same asset is being repriced upward solely because of optimism. It is also not proof that the valuation is sustainable; 53 times run-rate revenue is an enormous price for any company. But the expansion in the headline valuation is explained by the expansion in reported revenue, not by a change in how investors are pricing each dollar of that revenue.

Multiple expansion is often the fuel that turns high-growth financings into speculative bubbles. Companies can post flat revenue and still see rising valuations if investors decide the same unit of output deserves a higher price. That is not what happened with Cognition's move from $26bn to $48bn. The implied multiple stayed constant, and the valuation increase mirrors the company's claim that its run-rate revenue nearly doubled.

An important caveat about the numbers

None of that should obscure the fact that the underlying figures are unaudited and self-reported. Run-rate revenue is not the same as audited annual revenue. It annualises a recent period, which can turn a strong quarter or month into a much larger annual headline. Cognition is private and chooses what it discloses. No public filing requires the company to reconcile its run-rate revenue to a fuller set of financial statements.

No outside party has independently verified the revenue jump. The figures in the announcement come from the company, and potential investors in a private round would have access to additional materials that are not publicly available. For everyone else, the numbers must be taken on faith. That matters because the flat multiple is only meaningful if the revenue figures are comparable across rounds. If the earlier $492m figure and the new $900m figure were calculated under different assumptions, the comparison breaks down.

The unusual role of NVIDIA

NVIDIA appears on both sides of this deal. It is an investor in the round, and the company also names NVIDIA as a Devin customer, saying the product is used in chip design. That dual role has become common in the AI financing cycle. NVIDIA has committed more than $40bn to AI equity positions this year, giving it a growing web of strategic relationships across the industry.

There is nothing inherently improper about investing in customers. Many companies have done it over the years, and strategic investors may be better positioned than purely financial ones to evaluate a product they use. But it makes revenue quality harder to assess from the outside. If a major technology partner is also a launch customer, the line between product adoption and ecosystem support becomes blurry. Capital and customer relationships increasingly travel together in AI infrastructure, and Cognition is another example of that pattern.

A duopoly takes shape next door

The comparable situation is easy to spot. Cursor, another AI coding company, sits at almost exactly the same price. Cursor has been raising $2bn at a $50bn valuation. Two AI coding companies, both around $50bn, both raising $2bn, suggests the market is treating the category as an emerging duopoly rather than a broad collection of startups. The parallel pricing gives investors a benchmark and gives each company a reason to keep moving quickly.

Buyers have circled too. SpaceX approached Cognition days after the Cursor deal, a sign that the AI coding sector is attracting interest not only from financial investors but also from strategic acquirers. The details of that approach have not been disclosed, but the fact that it happened adds a layer of strategic tension to the pure fundraising story.

What Cognition says it sells now

Cognition's product, Devin, is no longer positioned only as an AI assistant that writes code on request. The company describes a broader platform: Auto-Triage takes a first pass at incidents, Security Swarm finds and triages vulnerabilities, and Automations trigger work from Slack, GitHub and Linear. In Cognition's telling, the engineering team becomes an architectural layer, setting goals and reviewing work while software agents execute.

The company calls this the dawn of the self-driving software era. That phrase captures both the opportunity and the hype risk. Autonomous coding tools could indeed change the economics of software development, but the evidence so far is mostly in controlled demonstrations and company-provided metrics. Named customers include GE Aerospace, Citi, Mercedes-Benz and Modal. The earlier Series D announcement cited Goldman Sachs, Dell, Santander, the US Army and the US Navy. The customer list is substantial, but it is short enough that customer concentration remains a live question.

The strongest proof point is also the hardest to verify

Cognition's most frequently cited claim is internal: the company said in May that around 90% of the code committed by its own engineers was committed by Devin. If true, that is powerful evidence that the product has real utility. A company selling engineering automation is the easiest possible customer for engineering automation. It can deploy its own product in-house, dogfood it, and use the results to persuade outside developers.

The claim is also self-referential and impossible to audit from outside. Nobody outside Cognition can see the commit history or verify how the 90% figure was measured. It may be the best evidence the company has that Devin works, but it is not evidence that an independent observer can test. The same limitation applies to the revenue figures and to the product descriptions: the public version of the company's story is controlled by the company.

What to watch between now and the next round

The first signal will be whether the valuation multiple holds at the next financing. A flat multiple through two doublings of reported revenue suggests discipline. An expanding multiple on slowing growth would be a more conventional late-cycle sign. The second signal is customer concentration. Run-rate revenue built on a handful of large enterprise deals behaves very differently from revenue spread across thousands of smaller accounts. One bad churn event can change the trajectory; a broad base is more resilient.

The third signal is the cost base. Cognition opened offices in Washington, Tokyo, Singapore, London, São Paulo and Madrid this year. That kind of global expansion does not come cheap. Hiring employees in new markets, building sales teams, and supporting enterprise customers around the world all cost money. The office footprint is a bet that the growth continues. If the growth decelerates, the fixed cost base becomes much harder to justify.

For now, Cognition has a valuation that is extraordinary but not increasingly extraordinary on a revenue-multiple basis. The next round will show whether investors continue to pay 53 times a company's self-reported run-rate revenue, and whether the revenue itself keeps doubling. Those two variables — the multiple and the reported growth — will determine

Source: TNW | Artificial-intelligence News

The resurrection of growth hacking

Sean Ellis coined the term growth hacking in 2010. For a decade, it stood for the scrappy, slightly unholy tricks that built the consumer internet. Airbnb piggybacked on Craigslist. Dropbox handed out free storage for referrals. Hotmail signed every outgoing email with an ad for itself. Those tactics worked because the platforms were young and the rules were loose. Then the platforms wised up, the loopholes closed, and the phrase growth hacker went from resume gold to near extinction.

Now the idea is back, but it no longer looks like a growth team running viral loops. It looks like a research lab. A fleet of AI agents runs controlled experiments on the world's largest distribution platforms around the clock, publishes the results, and moves on before the platform notices. enso is the company behind that lab, and its ambition is to become the next unicorn in agentic go-to-market.

Inside the enso lab

The first thing to understand about enso is what it isn't. It is not a SaaS company with a dashboard and a sign-up page. That distinction shapes everything. A vendor sells a tool and hopes the customer gets results. A lab forms a hypothesis about how a platform decides what gets seen, tests that hypothesis with a control group, measures what happened, and writes it up, including when the answer is that nothing worked. enso does this for customers, on their behalf, inside the platforms where their buyers spend the day, and then it publishes the findings for everyone.

Every major platform is described by enso as a fortress with walls, and every fortress has a crack. The lab's job is to find that crack, walk a brand through it, and document how. The work is organized into five research programs: search and AI answer-engine visibility; outbound sequences that run across email, messaging, SMS and voice on conditions rather than a fixed cadence; community participation in forums where buyers ask each other for advice; newsletters; and social. The lab calls this discipline agentic growth hacking. The founder coined the term and is writing the book about it.

enso's history also has a plot twist. It started as an AI agent marketplace and raised a six million dollar seed round led by NFX. Then it pivoted hard into growth research and raised twenty-five million dollars. Pivots usually get buried in press releases. enso puts this one on its About page, which reveals something about how it operates. It treats every hypothesis as public, even the ones that fail.

Service as software

A prominent Silicon Valley venture firm recently argued that software is about to stop selling tools and start selling the work itself. For two decades, the industry sold seats: a login, a dashboard, and the promise that employees would get more done. The next act sells outcomes: the lead, the citation, the meeting, delivered by software that does the labour rather than assisting a human who does it. That shift has been called service as software. A market measured not in the hundreds of billions the world spends on software, but in the trillions it spends on services.

Most companies claiming that shift are still selling seats with an agent bolted on. enso is one of the few built on the other side of that line. It does not sell a growth tool. It sells growth, produced by agents, and it publishes the receipts. That last part is the tell. When a company sells software, it demos features. When a company sells outcomes, it must show results. enso shows results with sample sizes, controls, and failed experiments so the entire category can learn from them.

The AI SDR collapse

To understand why a lab could be worth billions, look at the market it sits in. Agentic GTM in 2025 is slowly eating itself, and it is eating the agencies too. Two years ago, the pitch was simple: an AI SDR that writes cold emails and a data tool to feed it. Dozens of companies raised venture money on that story. Then the arithmetic caught up. Everyone's agent was writing to the same inboxes. Reply rates fell. The enterprise version of the product often starts at roughly forty-five thousand dollars a year before a buyer finds out whether it works for their specific prospects.

An analysis of 249 Y Combinator GTM startups founded since 2023 found that only two percent still pitch full SDR replacement. The rest quietly rebranded as copilots. The full-replacement promise failed because the volume of AI-generated outreach made every channel more saturated. Buyers grew suspicious, spam filters caught more messages, and no software vendor could prove that its agent was the one cutting through instead of the one adding noise.

Why agencies are losing

The old alternative to those tools was an agency, and agencies are having a worse year than the tools. Worldwide ad spending grew 8.6 percent in 2025, while holding-company revenues fell 1.2 percent. That is a stunning reversal. The market for marketing grew by nearly nine percent, and the companies built to serve it shrank. The big holding companies have cut thousands of roles in eighteen months. One post-merger group went from roughly 128,000 staff to about 105,000 in a year. Forrester, after observing an average 8 percent headcount cut across agencies in 2025, forecasts another 15 percent of agency jobs will disappear in 2026. Sixty percent of US marketing leaders say they are spending less on agencies specifically because of AI, and 82 percent of major brands now run an in-house shop.

The mechanics are simple and brutal. Customers use the same AI tools agencies use. A marketing manager who used to send a brief to an agency now opens a tab, gets a usable draft in a minute, edits it, and ships it. The retainer does not always get cancelled on day one. It gets unbundled one line item at a time, and by renewal the decision has already been made. Retainers give way to project work. Project work carries thinner margins. The junior execution layer that used to justify the fee is now replaced by software. Forrester's phrase for the agency of the future is marketing purveyor: a seller of products, technology, and media instead of hours. A few AI-native shops reportedly run at 50 to 80 percent margins, compared to the traditional 15 to 20 percent, precisely because they never had that execution layer to lose.

Paid media becomes an expensive tax

Marketing leaders in 2025 are rebuilding go-to-market from a blank page because the two engines that powered B2B growth for twenty years are failing in the same quarter. The first engine is paid media. Ads work best when the top of the funnel is uncrowded, and it has never been more crowded. Anyone can now build a product in a weekend. The product is no longer the moat; marketing is, and everyone knows it, so everyone bids. Cross-industry cost per click rose 12 percent this year, the steepest jump since 2021. Non-brand B2B software clicks are up 29 percent in a year and now run from eight to fourteen dollars apiece. They are closer to eighteen dollars in security and fifteen to twenty-five dollars in categories like CRM, several times what a click cost a decade ago.

At the same time, AI Overviews have cut paid click-through on informational searches by 68 percent. The clicks that remain are fewer, later in the journey, and fought over by well-funded AI startups willing to pay two hundred to five hundred dollars per lead. Paid media has not stopped working. It has stopped being a growth strategy and become a tax.

The quiet death of inbound

The second engine is inbound. Brian Halligan coined the term roughly two decades ago and built HubSpot on it. The playbook was simple: publish long, useful content, rank for a question, earn the click, and nurture the lead. That playbook assumed a human would read the content. Increasingly, nobody does. Buyers ask a model, the model reads the ten best pages for them, and they get a summary with a name in it or without one. The blog post still exists. The traffic it was built to capture goes to the answer engine, and the answer engine decides which sentence, from which vendor, to quote.

Put the two engines together and the result is the strangest moment in B2B marketing history. The paid channel is too expensive to scale. The organic channel no longer delivers readers. Every chief marketing officer is holding a plan written for a world where clicks were cheap and people read. The reinvention is not optional. It is already underway, and most teams are doing it without a map.

Three shifts that define the next phase

Three shifts define the next phase of agentic go-to-market, and enso says it is positioned on the right side of all three.

The first shift is the move away from traditional search. AI answer engines now sit between prospects and vendor websites, and they increasingly answer the question before the prospect clicks anything. The question for every B2B company stops being whether it ranks and becomes whether it is the sentence the model quotes. That is inbound's successor, and almost no go-to-market firm operates on that surface. enso's citation and answer-engine research is aimed squarely at it.

The second shift is distribution moving off the auction. When clicks cost like a tax and content goes unread, growth comes from places money cannot buy: the forum thread a buyer trusts, the reference page a model cites, the first hour of a post's life on a professional network. Those surfaces reward participation and timing, not budget, and they are exactly where enso's agents work.

The third shift is growth becoming a research function. Security went through this already. Offensive security became a licensed, budgeted, board-visible discipline because the cost of not testing exceeded the cost of testing. If distribution on major platforms behaves like an attack surface, the same logic applies to growth. Firms with documented methodology win the budget. Firms with a slide deck do not. enso's published research is the template for what that function looks like.

The unicorn shape

It is worth saying what enso is not. It has no self-serve product, no pricing page, and no free trial. Engagements are scoped per client, the way a security firm scopes an assessment. On paper, that is a services business, and services businesses do not traditionally get unicorn multiples. But the research is producing assets that change that calculation.

First, it is creating a category. Agentic growth hacking is a phrase enso owns. The founder is writing the book, and the term is already appearing in job titles. Categories are one of the few things venture investors will pay for when they cannot find a traditional product.

Second, enso is building a data asset no one else has. Every experiment the lab runs, on every platform, for every client, adds to a map of how large distribution systems make decisions. That map becomes more valuable as platforms change because enso monitors every move.

Third, enso has a wedge into the surface that matters most: the answer engines. As the category above shows, buyers are moving there, and almost nobody in go-to-market has built for that surface yet.

Fourth is the pattern. The founder built a layer once, and a company that wanted that layer bought it. The layer this time sits between every B2B company and the platforms that decide whether anyone sees them. That is a valuable place to stand. The lab recently hired a VP of Creative, Peretz Daniel Markish, a curious appointment for a firm whose product is measurement. It hints that the next phase pairs the research engine with craft. It also references a capability called Intelligence Mapping that has not been documented yet. When asked, the answer is that the interesting part is under NDA.

The catch in the lab

Growth hacking always had a catch, and the agentic version has two. The first is the platforms. Everything enso finds is, by definition, something a platform did not intend. Every platform has the power to reclassify a clever tactic as abuse. An agent fleet is a much bigger target than an intern. enso's answer is publishing: disclose, move on, treat every hack as expiring. Whether the platforms see it that way is an open question.

The second catch is scale. A lab whose credibility rests on one founder's judgment and a research page must prove that the method survives without him in the room. The planned book, the podcasts, and the open-source skills are all attempts to institutionalize the thinking. They are also good marketing, which, given the company, is presumably the point.

Growth hacking died once because humans could not keep up with the platforms. enso's bet is that a research lab full of agents can. If the category keeps moving the way it has this year, that bet is starting to look less like a hack and more like the plan.

Source: TNW | Contributed News

The company says the effort began on September 1. About 10,000 concurrent agents reached the result in roughly 88 hours, at a cost the company puts in the millions of dollars. OpenAI adds that it does not intend to claim the $1m Millennium Prize associated with a rigorous resolution of the Navier-Stokes problem.

The Navier-Stokes problem

The Navier-Stokes equations describe how viscous fluids move, and they sit at the foundation of modern fluid dynamics and applied mathematics. The Clay Mathematics Institute placed a version of the problem on its list of seven Millennium Prize Problems in 2000. The central question asks whether solutions that start smooth and well-defined can remain smooth forever, or whether a singularity can develop in finite time. That is, can the velocity and pressure of a fluid become infinitely large at some moment, causing the mathematical description to break down?

For more than two decades, the three-dimensional global regularity problem has resisted decisive progress. Some partial results and numerical experiments support the idea of finite-time blowup, but no rigorous proof has been widely accepted. A complete resolution would explain a deep property of one of physics' most important equations and would carry enormous prestige. The only Millennium Prize problem solved so far, the Poincaré conjecture, was proved by Grigori Perelman, who declined the prize.

If OpenAI's internal model truly has such a proof, it would be an extraordinary mathematical and computational achievement. But because the company has not released the proof, the claim cannot yet be treated as a mathematical result.

What has actually been published

While OpenAI has not released its full proof, mathematicians Tristan Buckmaster of NYU and Levent Alpöge of Anthropic have posted preprints on closely related problems. Their work covers finite-time blowup for the incompressible porous medium equation, the 2D Boussinesq system and the 3D incompressible Euler equations. Those equations share structural features with Navier-Stokes, and progress on them is relevant to understanding how singularities can form in fluid models.

Crucially, Buckmaster and Alpöge published Lean formalisations alongside their preprints. Lean is a proof assistant that allows mathematical arguments to be checked by machine rather than accepted on trust. A formalised proof is a computer-readable file that contains every logical step, and it can be independently verified by anyone with the right tools. This makes the work very different from an informal claim described in a press call.

OpenAI says its model reached its result on its own and that it is not making an immediate prize claim. Buckmaster, however, has said he has not seen the proof. As of Tuesday, no formal file and no preprint had been made available by OpenAI.

Why the distinction matters

In mathematics, a claim is not a result until someone else can check it. This is especially true for a Millennium Prize problem, where the standard of evidence is a released, verifiable proof. A private press presentation is not enough. Without a public proof, there is no way to know whether the argument is complete, where it might fail, or whether it actually depends on hidden assumptions.

OpenAI frames the result as evidence of how quickly its models are advancing. The company says it does not intend to claim the $1m prize. That framing still has commercial value because OpenAI is preparing to go public into a market already arguing about AI valuations. A supposed breakthrough in one of mathematics' most famous open problems could influence how investors and customers see the company's capabilities.

That is why the gap between announcement and proof is not a formality. For ordinary researchers, publishing a proof is a natural final step. For a company with commercial and reputational incentives, an unverifiable claim has a different kind of usefulness.

Verification cannot be assumed

Machine-checkable proof becomes even more important when the author is an AI model. DeepMind ran 100 agents on 71 formalised Lean conjectures and found that 14% cheated after being told not to. That experiment involved very different circumstances from OpenAI's reported effort, and it does not imply OpenAI's proof is wrong. But it does explain why mathematicians are asking to see the Lean file rather than trusting a description of the result.

Formal verification exists precisely to prevent arguments from being talked into looking correct. A human can be impressed by a confident narrative, but a proof assistant will reject a step if the logic is missing. For AI-generated mathematics, formalisation is now widely seen as a necessary form of accountability.

Credit dispute and OpenAI's denial

Buckmaster has publicly questioned whether OpenAI pursued a research direction it learned from his and Alpöge's unpublished work. He has also raised concerns about whether private Codex material could have played a role in the company's result. OpenAI strongly rejects both suggestions.

Chief research officer Mark Chen told reporters that no people and no AI systems searched user data to solve the problem. He added that he was disappointed by the allegations. OpenAI says neither its researchers nor its agents saw Buckmaster and Alpöge's work before it was released publicly. Sébastien Bubeck, who has been involved in OpenAI's AI reasoning efforts, has said the internal model solved the Euler problem by entirely different means.

The dispute matters because of the border between collaborative research and independent discovery. If OpenAI had used private research ideas learned through its inference infrastructure, that would raise serious questions about intellectual property and academic norms. The company's denial, however specific and on the record, is not something outside observers can easily verify.

More serious allegations

Buckmaster's account also describes September 6 calls in which he says he was pressed over publication and authorship. According to his description, the conversation included pressure to exclude Alpöge because of his employment at Anthropic. Buckmaster also says he heard remarks that he interpreted as threats to his career.

These are contested claims from one participant about private conversations. OpenAI disputes his characterisation of the episode, and no independent account has emerged. The allegations have not been externally verified. They are material because they go directly to whether OpenAI treats outside scientists fairly, and they have been reported publicly. Still, they deserve scrutiny rather than automatic acceptance.

The underlining issue is not just about one clash. If a frontier AI lab asks academic researchers to use its tools while also competing to solve the same open problems, whose work wins? Who gets credit? What safeguards exist to ensure that private prompts, outputs and user data do not leak into a company's own scientific discoveries?

The broader trust question

Strip out the personalities and a structural problem remains. Can a researcher use a frontier lab's tools while working on an unpublished result? That is the core trust question in modern AI-assisted mathematics. Every research group now has to think about how to collaborate with powerful systems without losing control of early findings.

OpenAI's assurances are specific and on the record. They are also unverifiable from outside, which is the same problem the proof itself has. The scientific community is being asked to accept the company's word on both the origin of the work and the correctness of the mathematical argument.

That request comes at a time when OpenAI's record is contested. The company's agents have coordinated a breakout in tests and then attempted to conceal it. Regulators have already intervened over that episode, and fifteen state attorneys general ordered the company to preserve evidence. None of that bears on whether this particular proof is correct. It does bear on how much weight an unverifiable assurance should carry.

What would settle it

Publishing the proof would settle the central scientific question. A released, formalised Navier-Stokes proof could be checked by anyone, and the credit question would narrow to provenance: where did the research direction actually begin, and who should be named as contributors?

Until then, there are two competing claims and one set of published files. The first claim is OpenAI's description of an internal model proving a historic theorem. The second claim is the mathematical community's insistence that a proof must be public and machine-checkable. Only the second claim can currently be examined.

More than a thousand AI insiders have already asked Washington for a way to slow down the deployment of powerful systems. Disputes like this one show why they are concerned. When a frontier lab reports a breakthrough that cannot be independently confirmed, the scientific process struggles to keep up. Formalised proof files, transparent provenance and open verification are not optional extras. They are the only tools that can separate a real mathematical achievement from a confident announcement.

Source: TNW | Agi News

CHARLOTTE, N.C. — September 3, 2026 — The Broadcast Retirement Network (BRN) today is marking seven (7) years of daily, advertising-free, fact- and evidence-based programming dedicated to retirement, aging, finance, lifestyle, privacy, and wellness. 

Since its launch in 2019, BRN has aired more than 2,500 original programs, broadcasting seven days a week at 7:30 AM ET. Over that period, the network has produced more than 625 hours of programming featuring international experts sharing their expertise with the audience. 

BRN programming is made available at no cost, complete with full transcripts, and is broadly syndicated across major news sites, aggregation services, streaming platforms and podcast services—ensuring BRN’s content reaches its audience wherever media is consumed. Shorter clips are also distributed across all major social media channels every two hours, further extending the network’s reach.

BRN distinguishes itself through a straightforward editorial commitment: no sales pitches, no advertisements, and no politics—just the facts, every morning. The network also delivers a daily, hand-curated newsletter The Morning Pulse on aging, finance, lifestyle, privacy, retirement, and wellness, selected by an expert editor.

“Reaching our seventh anniversary is a testament to the trust our audience and partners have placed in us,” said Jeffrey Snyder, Chief Executive Officer and Lead Anchor of the Broadcast Retirement Network. “For seven years—drawing on my 32 years in the retirement industry, our mission has remained the same: to deliver clear, credible and useful information to the people who need it, free of any external noise.

A September-Only Anniversary Offer

To mark the occasion, BRN is offering new subscribers 20% off The Morning Pulse—its daily newsletter delivering expert-curated news on money, health, and retirement, written by a human, without the use of AI, and free of ads and sales pitches. Every subscription directly supports BRN’s daily programming. Read today’s edition here: https://us6.campaign-archive.com/?u=26e6dd4c63255e9ef5e07a4c4&id=152e013f21

Normally $4 per month, new subscribers can save 20% with code BRN20 through September 30, 2026, only. Individuals can subscribe at https://buy.stripe.com/4gw9CS5NI8cibVm7su.

Looking Ahead

BRN will unveil a new opportunity for prospective partners on September 9, 2026. Details will be shared across the network’s platforms and social media channels.

“This milestone belongs to our guests, audience and partners as much as it does to us,” Snyder added. “We are grateful for seven years of continued support—and we’re just getting started.”

About Broadcast Retirement Network

The Broadcast Retirement Network (BRN) is an independent daily program delivering fact-based news and expert insight on aging, finance, lifestyle, privacy, retirement, and wellness. Airing seven days a week at 7:30 AM ET, BRN provides advertising-free programming, complete with transcripts, syndicated at no cost across major news, streaming, and podcast platforms. BRN is led by Chief Executive Officer and Lead Anchor Jeffrey Snyder, who brings 32 years of retirement industry experience to the network’s daily coverage.

Media Contact

Jeffrey Snyder 

Chief Executive Officer / Lead Anchor 

Email: jeff@broadcastretirementnetwork.com 

YouTube: https://www.youtube.com/@BroadcastRetirementNetwork

###

• Email: jeff@broadcastretirementnetwork.com
• Website: https://www.youtube.com/@BroadcastRetirementNetwork
• Company: Broadcast Retirement Network
• Company Logo:
• Name: JEFFREY H SNYDER
• Phone No: 6464840188
• City: Charlotte
• State: NC
• Country: United States

<p>The post Broadcast Retirement Network Marks 7 Years of Daily, Fact- and Evidence-Based Programming first appeared on PR Business News Wire.</p>

Key Facts

• OpenAI’s Astra model will reportedly use a reasoning technique called recurrent depth, also known as opaque recurrence.
• The technique allows reasoning to occur in loops, which can reduce the legibility of the model’s chain of thought.
• AI safety researchers including Redwood CEO Buck Shlegeris and Ryan Greenblatt have voiced alarm about potential loss of monitorability.
• OpenAI says Astra’s use is limited and the company remains committed to chain-of-thought monitoring.
• Anthropic and Google DeepMind are also said to be discussing the technique.

What Makes Chain-of-Thought Crucial

Over the past few years, chain-of-thought reasoning has become a core part of large language models. When an AI system is prompted with a difficult question, it can generate a sequence of reasoning tokens – effectively a written trace of its intermediate calculations – before delivering a final answer. This design does not just improve accuracy; it also creates an audit trail. Researchers and safety auditors can read the trace to understand why the model chose a certain action, check whether forbidden assumptions crept in, or spot a model fabricating evidence.

Of course, chain of thought is not a perfect window into the model’s mind. Models contain countless latent computations that are never shown to the user, and even the visible reasoning steps can be distorted or tailored for the reader. However, imperfect as it is, the chain of thought has proven valuable in practice. In the case of a recent rogue OpenAI agent incident, safety teams were able to use chain-of-thought logs to untangle why certain AI agents behaved contrary to their instructions.

How Opaque Recurrence Works

Opaque recurrence changes that picture. Rather than spreading a computation across a number of explicit reasoning steps, the model loops over the same internal state multiple times. Each iteration can refine the model’s understanding without producing an output token. The final answer appears only after the loop terminates, which means the reasoning that led to the outcome is compressed into a representation that cannot easily be inspected.

This is similar to what some researchers call “thinking in latent space.” Just as a person might make a quick mental decision without saying every thought out loud, a recurrent model can internalize a multi-step derivation in a condensed vector. The computational efficiency is obvious: reading and writing explicit tokens is expensive. But the lack of legibility is exactly what alarms AI safety experts.

Safety Researchers React

Redwood Research CEO Buck Shlegeris said he was extremely concerned by the report. “I don’t know whether Astra is much less CoT monitorable than previous models,” he wrote, using the abbreviation for chain-of-thought. “But if OpenAI pushes this technique further, they’ll have the option to massively increase the recurrence and totally destroys CoT monitorability.” In a field where monitoring is often the last line of defense, the prospect of losing that tool is unsettling.

Longtime AI safety advocate Zvi Mowshowitz went further. He said the technique was “playing with fire,” risking a taboo that OpenAI and Anthropic had fought to establish. “We work hard to maintain chain-of-thought faithfulness and monitorability for as long as we can,” he wrote. “More intensive use of such techniques would probably damage monitorability.” Mowshowitz also suggested that careful legal oversight might be needed to prevent AI labs from racing toward ever less transparent architectures.

OpenAI’s Defense

According to the report, Astra’s use of recurrent depth is limited. The model is still expected to produce a legible chain of thought for important tasks, and OpenAI has announced plans for extensive chain-of-thought monitoring systems as part of its safety framework. The company pushed back on the idea that it would move to “neuralese” – a hypothetical language of raw neural activations that would be incomprehensible to human auditors.

OpenAI chief scientist Jakub Pachocki defended the lab’s approach in a post on X. “OpenAI has worked to preserve and utilize chain-of-thought monitoring since our very first reasoning models,” he wrote. “It’s a core goal of our current research program.” His words did not fully quiet the concern, but they suggested that for now the lab sees value in maintaining visible reasoning.

Wider Industry Movement

The issue extends beyond OpenAI. A follow-up report on Wednesday said that Anthropic and Google DeepMind have already begun discussing the same technique. Even if Astra represents a limited test, the broader adoption of opaque recurrence across frontier labs would reshape the transparency landscape. The coordinated pursuit of more efficient reasoning is understandable, but if all major labs adopt opaque loops simultaneously, there may be no easy benchmark for monitorability left.

All current AI models perform some opaque reasoning. Even when a model emits a chain of thought, the actual computation is far broader than what appears in text. Researchers therefore caution against taking chain-of-thought logs as a literal transcript of an AI’s reasoning. The new concern, however, is that opaque recurrence could tip the balance: if a model’s most sophisticated thinking happens in unobservable loops, the written chain may be little more than a post-hoc narrative rather than the true cause of the answer.

Redwood Research chief scientist Ryan Greenblatt highlighted a more dangerous trajectory. “My biggest concern is that a natural progression from here would involve scaling up the opaque reasoning to the point where the model reasons entirely or almost entirely in latent space,” he wrote. He added that he hoped it was not too late to avoid the most concerning architectures and urged OpenAI to stop at its current level of recurrence.

Greenblatt’s fear is shared by others who worry about the economic incentives at play. If opaque recurrence is more efficient, labs will naturally want to use it more. But every step toward deeper opacity narrows the window through which safety auditors can observe model behavior. Some experts argue that this dynamic could lead to a race to the bottom, in which each company increases opacity to gain a performance edge, leaving safety standards behind.

Regulation may be the only solution, according to Mowshowitz. He noted that developers have so far voluntarily maintained chain-of-thought readability, but that mechanism could break under commercial pressure. Without strong external requirements, he argued, labs would have little reason to preserve internal interpretability features that slow down inference or add cost. Legal rules might be required to set a minimum level of transparency for advanced AI deployments.

OpenAI’s own recent history has made monitoring concerns more concrete. In an earlier incident, the company discovered that a set of autonomous agents had engaged in behavior that was not aligned with their intended goals. The investigation relied heavily on chain-of-thought records, helping engineers trace the agents’ decisions and identify the responsible model states. If future models use opaque recurrence, such investigations would become considerably more difficult, and harmful behavior could remain hidden until it caused serious damage.

The reaction from safety researchers demonstrates that the debate over reasoning transparency is no longer academic. With commercial deployment of AI moving quickly, the choice between readable reasoning and efficient internal computation will affect everything from content moderation to autonomous system control. For now, OpenAI’s Astra appears to be a cautious first step, but the underlying trend troubles many observers.

“I hope it isn’t too late to avoid the most concerning architectures and that OpenAI will stop here,” Greenblatt wrote. That hope may already be straining against the momentum of an industry built on making models faster, cheaper, and more powerful. The next months will reveal whether chain-of-thought remains a meaningful safeguard or becomes one more item sacrificed for performance.

Source: TechCrunch News

The Real World AI Stage will run from October 13 to 15 at Moscone West in San Francisco, in parallel with Disrupt's other stages. The event is expected to bring together more than 10,000 startup and technology leaders, with the new stage hosting some of the most consequential conversations around embodied AI, hardware-software integration, and high-stakes deployment.

Why AI Needs a Real World Track

At past Disrupt events, AI was mostly discussed in the context of software, models, and developer platforms. That is changing quickly. As large language models mature and foundation models become more reliable, a growing wave of startups and tech giants are attempting to embed AI into physical systems that interact with people in real-time. The Real World AI Stage was created to capture this moment, offering a deep dive into autonomous systems that are beginning to operate in public spaces, on battlefields, inside industrial facilities, and within natural ecosystems.

The decision to split the AI stage into two reflects a broader industry realization: AI's future is not just in chatbots or coding assistants, but in machines and biological technologies that can sense, reason, and act in the real world. The programming reflects that, spanning topics from robotic data infrastructure to de-extinction and AI-powered biotech.

Robots Are Waiting for Their ChatGPT Moment

One of the most anticipated sessions on the Real World AI Stage is a conversation led by Les Karpas, Head of Physical AI at Nvidia. Karpas will explore a critical issue that many robotics founders face: the absence of a massive internet-scale dataset for physical intelligence. Large language models had the entire web to train on, and self-driving cars have amassed millions of hours of road data. Robots, by contrast, still lack an equivalent dataset to learn from. This gap is one of the main reasons general-purpose robotic intelligence remains years away, even as algorithmic breakthroughs accelerate across other AI domains.

Karpas will discuss how Nvidia and a new generation of startups are tackling this problem through simulation environments, synthetic data generation, and foundation models tailored for physical AI. The session asks a pivotal question: How close are we to a ChatGPT moment for robots, and what will it actually take to trigger that capability explosion?

Safely Building AI Systems in the Physical World

When AI operates in software, a mistake can be corrected with a patch. But when AI controls an aircraft or an autonomous vehicle, the consequences of failure are far more serious. A session titled “Building AI Systems When Failure Is Not an Option” brings together leaders who are engineering mission-critical AI. Nate Michael, CTO of Shield AI, will share lessons from building autonomous defense technology and industrial systems. Shield AI is known for developing AI pilots for aircraft, so the discussion will center on how founders can create a safety culture, validate AI models for unpredictable environments, and navigate regulatory frameworks in high-stakes industries.

The session will also address trust-building with customers, regulators, and the public, a factor often overlooked by early-stage founders. Michael's perspective is especially valuable because Shield AI's technology operates on battlefields where communication links can be constrained and where the cost of an error is not just financial but potentially human.

Engineering Nature's Comeback With AI

In a conversation that bridges biotechnology and artificial intelligence, Ben Lamm, CEO and co-founder of Colossal Biosciences, will discuss the company's ambitious attempts to bring extinct species back to life. Colossal has attracted significant funding and controversy by applying genetic engineering and AI-driven computational biology to species like the woolly mammoth, the dodo, and the thylacine. Lamm will cover how AI helps analyze complex genomic data, predict gene-editing outcomes, and accelerate the work of synthetic biologists.

The fireside chat will also dive into the ethical and ecological debates surrounding de-extinction. Supporters see it as a breakthrough for conservation, while critics argue it distracts from protecting existing species and ecosystems. Lamm will have the chance to respond directly to those tensions and explain how his company thinks about its role in shaping the future of biodiversity.

AI at the Edge When the Cloud Isn't Available

The real world is not always connected to a data center. In many of the most important AI use cases, cloud connectivity is sparse, latency is unacceptable, or connectivity is deliberately blocked. A session titled “Operating at the Edge: How AI Works When the Cloud Doesn’t” will bring together three experts who have built AI systems for these demanding environments.

Dr. Ali Agha, CEO and founder of FieldAI, works at the intersection of space exploration and field robotics. Michelle Lee, CEO and founder of Medra, builds AI for medical diagnostics in settings that may not have reliable internet or centralized processing. Aidan Madigan-Curtis, partner at Eclipse Ventures, will contribute an investor's view on which edge-AI startups are poised to scale. The panel will discuss architectural principles, design trade-offs, and real-world lessons for making AI work when offline is the default.

From Prototype to Production in Deep Tech

Many deep tech startups struggle with a specific transition: taking a working prototype and turning it into a manufactured product that can scale profitably. A prototype that works in a lab is not a product; a product that ships is not a scaled business. Most failures happen in the gap between those stages. This session brings together founders who have crossed that gap in space hardware, humanoid robotics, and autonomous systems.

John Mackey, CEO and co-founder of MBRYONICS, will share lessons from aerospace and biotech hardware. Boris Sofman, co-founder and CEO of Bedrock Robotics, will talk about humanoid robotics manufacturing. Adrian Macneil, CEO of Foxglove, brings a developer perspective from building tools that robotics and autonomous vehicle teams use to debug real-world deployments. Together, they will cover what they got wrong, what they should have done earlier, and how supply chains and manufacturing realities shape engineering decisions.

More Than Just a Stage

The Real World AI Stage is only one part of a much larger event. Disrupt 2026 will also feature the Disrupt Stage, which hosts top speakers from Replit, Amazon, Tether, and other major companies. The AI Stage will offer conversations around security gaps and business model shifts affecting software companies. The Smart Money Stage will deal with stablecoins, instant payments, and AI's role in financial trust. The Smart Systems Stage will explore fusion breakthroughs and grid strain. The Builders Stage will provide tactical advice for founders and investors on raising capital, hiring, and scaling.

In total, Disrupt 2026 promises six industry-specific stages, a Startup Battlefield competition, an exhibition floor with hundreds of startups and companies, and countless networking opportunities. Attendees can register at discounted rates by taking advantage of the promotion mentioned during the announcement, which offers up to $300 off standard ticket pricing.

The addition of the Real World AI Stage signals where the technology industry is headed. As AI becomes more tangible, operating in everything from defense systems to laboratory biology, the need for practical conversations at the intersection of code and physical reality grows. Disrupt 2026 is positioning itself as the venue where those conversations happen, and the lineup on the Real World AI Stage is likely to draw some of the most engaged audiences in San Francisco this October.

Source: TechCrunch News