Charles Dickens’ miserly mister becomes a devilish fashion designer in this marvellously modern production

German magazine Der Spiegel has reported the warning, quoting a leaked note from a recent call between the European leaders

But on a more serious note, we are hearing that German chancellor Friedrich Merz has postponed his planned visit to Norway scheduled for Friday and will travel to Brussels instead.

Merz will travel to Belgium for a private dinner with Belgian prime minister Bart De Wever and European Commission President Ursula von der Leyen, his spokesperson said in comments reported by Reuters.

The tour kicks off in January at Birmingham’s Utilita Arena

Former local government minister Jim McMahon has hit out at the decision to postpone mayoral elections after Reform compared the PM to a dictator

Admiral Frank “Mitch” Bradley - described as ‘a leader of courage, integrity, professionalism and unmatched skill’ - will address Congress over the deadly ‘double tap’ strike on a suspected drug boat

NYT accuses Pentagon of infringing on the constitutional rights of its journalists

A new Emerson College Polling/Inside California Politics survey takes a look into the crowded race for California governorship, in which more than 95 candidates have so far submitted paperwork indicating their intention to run.

The poll found that Republican Chad Bianco, Republican Steve Hilton, Democrat Eric Swalwell (12%) and Democrat Katie Porter are currently leading the June 2026 primary. Of those polled, 13% said they were voting for Bianco, 12% for Hilton, 12% for Swalwell and 11% for Porter. In all, 31% of voters are undecided.

Wero poursuit son expansion. Après son intégration à Revolut, le service de paiement entre particuliers va se faire une place au sein de N26. À partir de l’été 2026, les clients français, allemands et néerlandais de la banque mobile pourront envoyer de l’argent à d’autres utilisateurs Wero depuis l’app N26.

La néobanque permet déjà de transférer facilement de l’argent à un autre client N26. L’intérêt de Wero est d’étendre cette simplicité aux utilisateurs d’autres établissements. Il n’y a pas d’IBAN à saisir, un numéro de téléphone ou une adresse email suffit pour effectuer le transfert.

« Notre solution de paiement sera désormais accessible à des millions d’utilisateurs supplémentaires de cette banque numérique sur les marchés où Wero est déjà opérationnel. Nous avons hâte qu’ils découvrent notre solution de paiement européenne sécurisée, fluide et rapide, notamment à l’heure où l’autonomie numérique de l’Europe est un enjeu de plus en plus central », déclare Martina Weimert, la CEO d’EPI, le consortium pilotant le service.

Wero est déjà intégré à la plupart des applications des banques françaises, puisque celles-ci ont poussé le projet avec leurs homologues belges et allemands. En France, le service permet uniquement de réaliser des transferts d’argent entre particuliers à l’heure actuelle, mais son périmètre va s’élargir. En Allemagne, il permet depuis peu de payer sur certains sites marchands, une fonctionnalité qui sera déployée dans l’Hexagone en 2026. À terme, il permettra aussi de régler ses achats dans les boutiques physiques, ce qui le mettra directement en concurrence avec Apple Pay.

Wero fourbit ses armes face à Apple Pay

N26 propose quatre offres bancaires pour les particuliers, qui vont de 0 € à 16,90 €/mois. La formule gratuite s’est nettement appauvrie au fil des ans, elle ne comprend même plus de carte physique. Mais dans le même temps, la néobanque a élargi son catalogue de services, en proposant notamment des produits d’épargne et d’investissement.

Lord Evans of Watford and other directors of investment firm deny claims made in lawsuits they say are ‘meritless’

A peer suspended by the House of Lords for breaking lobbying rules is now facing claims that he received at least $1m (£760,000) from an allegedly corrupt deal.

Lord Evans of Watford, a longtime Labour peer, was found last week by the House of Lords watchdog to have broken its rules four times after undercover reporting by the Guardian, and will be suspended for five months.

President orders pause on asylum claims and extra scrutiny of migrants from 19 countries after Washington shooting

Donald Trump is seizing on the shooting of two national guard members, allegedly by an Afghan man, to press his immigration crackdown still farther. In the aftermath of the attack, which left guard member Sarah Beckstrom dead and colleague Andrew Wolfe in critical condition, Trump directed US Citizenship and Immigration Services to pause all pending asylum applications.

USCIS followed up that announcement with more seismic shifts to immigration policy. This is how the White House is reshaping the process for requesting asylum, green cards and citizenship.

From income to home ownership and public services, everybody is disgruntled – but for different reasons

The British economy has endured a series of setbacks in recent years: austerity, Brexit, the global pandemic, soaring energy prices and an increasingly fractured and uncertain world. The early optimism that accompanied Labour’s election victory faded quickly, with a recent poll ranking Rachel Reeves as the worst chancellor in modern history.

But is this fair? Is the economy really in as bad a shape as people feel?

Jonathan Swarbrick is a lecturer in economics at the University of St Andrews Business School

Cyril Hanouna animait une nouvelle émission de Tout beau, tout n9uf ce mercredi 3 décembre 2025 sur W9. Durant l'émission, l'animateur a renforcé son soutien à Matthieu Delormeau d'une façon inattendue.

The shape and structure of blood cells provide vital indicators for diagnosis and management of blood disease and disorders. Recognizing subtle differences in the appearance of cells under a microscope, however, requires the skills of experts with years of training, motivating researchers to investigate whether artificial intelligence (AI) could help automate this onerous task. A UK-led research team has now developed a generative AI-based model, known as CytoDiffusion, that characterizes blood cell morphology with greater accuracy and reliability than human experts.

Conventional discriminative machine learning models can match human performance at classifying cells in blood samples into predefined classes. But discriminative models, which learn to recognise cell images based on expert labels, struggle with never-before-seen cell types and images from differing microscopes and staining techniques.

To address these shortfalls, the team – headed up at the University of Cambridge, University College London and Queen Mary University of London – created CytoDiffusion around a diffusion-based generative AI classifier. Rather than just learning to separate cell categories, CytoDiffusion models the full range of blood cell morphologies to provide accurate classification with robust anomaly detection.

“Our approach is motivated by the desire to achieve a model with superhuman fidelity, flexibility and metacognitive awareness that can capture the distribution of all possible morphological appearances,” the researchers write.

Authenticity and accuracy

For AI-based analysis to be adopted in the clinic, it’s essential that users trust a model’s learned representations. To assess whether CytoDiffusion could effectively capture the distribution of blood cell images, the team used it to generate synthetic blood cell images. Analysis by experienced haematologists revealed that these synthetic images were near-indistinguishable from genuine images, showing that CytoDiffusion genuinely learns the morphological distribution of blood cells rather than using artefactual shortcuts.

The researchers used multiple datasets to develop and evaluate their diffusion classifier, including CytoData, a custom dataset containing more than half a million anonymized cell images from almost 3000 blood smear slides. In standard classification tasks across these datasets, CytoDiffusion achieved state-of-the-art performance, matching or exceeding the capabilities of traditional discriminative models.

Effective diagnosis from blood smear samples also requires the ability to detect rare or previously unseen cell types. The researchers evaluated CytoDiffusion’s ability to detect blast cells (immature blood cells) in the test datasets. Blast cells are associated with blood malignancies such as leukaemia, and high detection sensitivity is essential to minimize false negatives.

In one dataset, CytoDiffusion detected blast cells with sensitivity and specificity of 0.905 and 0.962, respectively. In contrast, a discriminative model exhibited a poor sensitivity of 0.281. In datasets with erythroblasts as the abnormal cells, CytoDiffusion again outperformed the discriminative model, demonstrating that it can detect abnormal cell types not present in its training data, with the high sensitivity required for clinical applications.

Robust model

It’s important that a classification model is robust to different imaging conditions and can function with sparse training data, as commonly found in clinical applications. When trained and tested on diverse image datasets (different hospitals, microscopes and staining procedures), CytoDiffusion achieved state-of-the-art accuracy in all cases. Likewise, after training on limited subsets of 10, 20 and 50 images per class, CytoDiffusion consistently outperformed discriminative models, particularly in the most data-scarce conditions.

Another essential feature of clinical classification tasks, whether performed by a human or an algorithm, is knowing the uncertainty in the final decision. The researchers developed a framework for evaluating uncertainty and showed that CytoDiffusion produced superior uncertainty estimates to human experts. With uncertainty quantified, cases with high certainty could be processed automatically, with uncertain cases flagged for human review.

“When we tested its accuracy, the system was slightly better than humans,” says first author Simon Deltadahl from the University of Cambridge in a press statement. “But where it really stood out was in knowing when it was uncertain. Our model would never say it was certain and then be wrong, but that is something that humans sometimes do.”

Finally, the team demonstrated CytoDiffusion’s ability to create heat maps highlighting regions that would need to change for an image to be reclassified. This feature provides insight into the model’s decision-making process and shows that it understands subtle differences between similar cell types. Such transparency is essential for clinical deployment of AI, making models more trustworthy as practitioners can verify that classifications are based on legitimate morphological features.

“The true value of healthcare AI lies not in approximating human expertise at lower cost, but in enabling greater diagnostic, prognostic and prescriptive power than either experts or simple statistical models can achieve,” adds co-senior author Parashkev Nachev from University College London.

CytoDiffusion is described in Nature Machine Intelligence.

The post Generative AI model detects blood cell abnormalities appeared first on Physics World.