Longtime Slashdot reader schwit1 shares an opinion piece from the Washington Post, written by Virginia Governor Abigail Spanberger: NextEra Energy, a Florida-based utility company, has filed paperwork to buy Dominion Energy for about $67 billion, creating the largest regulated electric utility in the world. Virginia's State Corporation Commission is the regulatory body tasked with reviewing the application, and the SCC's commissioners will ultimately decide whether to approve, deny or impose new conditions on any potential merger. As a Virginian, I am deeply skeptical about whether selling our primary state-regulated utility to an out-of-state company is good for the commonwealth. I have serious questions about what this deal would mean for us. And as governor, I intend to get answers and be a voice for Virginians in the process. That is why I will be taking the legal step of "intervening" in this proposed merger, which means that as governor, I will formally request to be a party to the case. I know this action is unprecedented by a Virginia governor -- but so, too, is the size of this proposed merger and its potential impact on the commonwealth. Virginians deserve to know that their leaders are laser-focused on ensuring that their needs are part of the SCC review. "If two large corporations stand to benefit financially from this merger, so, too, should the Virginians who pay the bills," said Spanberger. "That is why any potential deal must deliver a more affordable energy bill with sustained, long-term energy cost savings." By formally intervening in the merger review, Spanberger's administration would gain legal standing to participate directly in the case, request detailed information, raise concerns, and advocate for conditions that benefit people in the state. She says the goal is to push for lower long-term energy costs, protect utility jobs, and ensure any new owner continues investing in reliable, affordable and cleaner power.

Read more of this story at Slashdot.

The National Rural Water Association has partnered with DEF CON Franklin to launch the Water Watch Center, which will provide threat intelligence and cybersecurity services to smaller U.S. water utilities serving fewer than 10,000 people. The initiative comes amid a growing wave of cyberattacks on water systems across multiple states. The Record reports: The program will see five managed detection and response providers work with DEF CON Franklin and the NRWA to offer cyber services for water utilities serving fewer than 10,000 people. There are over 50,000 community water systems across the U.S. and 91% of them serve fewer than 10,000 people. "These leading cyber firms and NRWA are architecting a scalable cyber delivery model that has eluded water industry and national security officials to date," said Jake Braun, the co-founder of DEF CON Franklin and former cyber official in the Biden administration. "Our incredible Franklin volunteers will support the effort to keep costs down for cash-strapped utilities fending off adversaries like the Iranian Red Guard or Chinese Military," Braun said. "To further accelerate the effort, we hand-picked cyber providers who already support water utilities. These folks walk in the door knowing the intricacies of cybersecurity in a water utility environment." The providers include Rapid7, Defendify, Legato Security, L1 Secure and Sentinel Technologies. The companies will share threat information and data on vulnerability patches with the NRWA, which will operate as a hub.

Read more of this story at Slashdot.

An anonymous reader quotes a report from NPR: Billions of dollars are traded every week on the lightly regulated prediction market sites, where users bet on everything from movie reviews to elections to conflicts in the Middle East. Clinical trials are just the latest area where the industry's rapid growth is raising ethical questions. Kalshi claims such bets will provide a new source of information about which drugs will get approved, and what clinical trials will show promising results, which the company says can help investors decide what new drugs to fund. "If you want to ban profiting from the failure of clinical trials, you would start with the stock market, where the financial incentive for this type of profit is orders of magnitude larger," said Kalshi spokesman Jack Such, pointing to stock market short sellers who have profited from clinical trial failures. "While Kalshi and the stock market are the same in this regard, they do differ in one important way: the stock market doesn't give any valuable information to researchers," Such said. Drug trial researchers, though, are far from convinced. David Tsai, who runs clinical trials at a biotech company in the San Francisco Bay Area, started an online petition pushing for such betting to be banned, making the case that betting on drug trials "threatens the very foundation of trust and integrity in biotechnology." Tsai is concerned that the prospect of betting provides those involved with a clinical trial a reason to tamper with the results for a prediction market payout. "If we were running a trial for an oncology drug that requires an infusion, a pharmacist who had placed a bet saying that it's gonna work well, or doesn't work well, could obviously adjust the infusion rate, could adjust the source temperature of the drug," he said. "They could change any number of variables that could obviously have a direct impact [on] how the trial and the data and the patient safety would come out." Another skeptic is Nicholas Zaorsky, a professor of radiation oncology at the Mayo Clinic in Jacksonville, Fla., who has helped run clinical trials and agrees that prediction markets can interfere with the advancement of life-saving drugs. "Prediction markets can be valuable in some settings because they aggregate information, but clinical trials are fundamentally different: investigators, coordinators, and sometimes even participants can directly influence aspects of the outcomes being wagered on," Zaorsky said. "That creates financial incentives that risk undermining trial integrity." Bettors should not be rooting for an experimental medicine to fail just to earn a buck, says Joshua Pederson, the father of a 12-year-old cancer patient enrolled in a clinical trial. "It's a dark idea," he said. "It's quite ghastly." Kalshi, for its part, argues that its prediction markets could help patients track promising medical breakthroughs and clinical trials, enlisting experts including 23andMe founder Anne Wojcicki to make the case. "Most patients don't know about the choices available in clinical trials or which programs are most promising. The opportunity to have an open, transparent dataset about trial probabilities is extremely promising and empowering for people," a white paper sponsored by Kalshi stated.

Read more of this story at Slashdot.

Ligue1+ sur Bbox diffuse l’intégralité de la Ligue 1 McDonald’s, avec 9 matchs en direct chaque journée. La compétition reprend le 23 août 2026, et vous y accédez depuis votre décodeur ou l’appli b.tv. Qu’est-ce que Ligue1+ ? Calendrier de la reprise Prix sur Bbox Comment y accéder Questions fréquentes 9 Matchs Ligue 1 en […]

L'article Ligue1+ sur Bbox : prix, chaîne et calendrier 2026 a été publié en premier sur Bbox-Mag

Former U.S. National Cyber Director Chris Inglis says the biggest AI risk isn't sentience but autonomy. "What I'm worried about is that they get to choose what and where they do something, and under what rules they do it," he said, citing recent cases of AI agents from OpenAI, Anthropic, and Meta escaping security sandboxes. He argues developers need stronger safeguards, monitoring, and human accountability, invoking Asimov's idea that protecting humans should come before simply obeying them. The Register reports: "Asimov was right," he said, referring to science fiction author Isaac Asimov and his three laws that were to be followed by robots -- more specifically, AIs, in this case. "The first rule, and we call it the superior role, must be that it's designed not to hurt humans," Inglis said. "Second rule: To obey humans, such that it doesn't achieve agency and aspiration on its own. And the third: To do what humans tell it - and in that order. Instead we've designed them in the exact opposite way." What this means, he explained, is that AI developers created models to "do what humans tell you, obey the humans until it's inconvenient, and then the third one is maybe implied - protect humans - but if that's not built into the DNA, hardwired into it, then we have no right to expect it." Inglis admits it's not possible to hardwire rules into models and still keep their non-deterministic nature. "I would offer that you can tease those out in a highly controlled environment, a true sandbox, where you say, 'Let's put this thing through its paces, and let's back away to see what happens,'" he said. "Maybe you get the equivalent of a mini nuclear explosion in that room, and now you know this thing is capable of that." Inglis thinks another problem with AI is that it's become a commodity. "It's not like you can control it like you can nuclear material," he said. "You can't even specify its properties the way you can for an airplane or for an automobile, as diverse as they might be. Its manifestations are so numerous, so diverse, that as a general matter, you can't actually win by simply saying, "I will design those properties in,'" he added. "You need to do that to some degree, and then make sure that you understand how to watch it, monitor it, make sure you know what it does." [...] Ultimately, humans remain accountable for AI models' actions, according to Inglis. "They remain the source of agency and aspiration. It's possible for them to give broad authority to an AI model and have it run around for 30 hours without further consultation, but they need to know what they've asked it to do, and they need to know what they expect it will deliver in terms of performance on the back end. If they don't, then they're going to get what they deserve, which is the very frequent unpleasant surprise."

Read more of this story at Slashdot.

For those with newer, high-end Logitech input devices, coming with Linux 7.3 is support for the re-programmable button support of the HID++ 2.0 protocol...

Upstreamed for the Linux 7.2 kernel was the Device Tree (DT) support for booting the mainline Linux kernel on the base Apple M3 SoC. Though it's not yet useful for end-users with many features remaining to be supported and upstreamed. Now for the upcoming Linux 7.3 kernel, the M3 Pro, Max, and Ultra variants will also have initial support for booting on the mainline kernel...

"Framework has been sending out email notifications to customers alerting of a limited data breach in which customer information was accessed through a Metabase BI service zero-day exploit," writes Slashdot reader DuoDreamer. Data includes customer names, email addresses, phone numbers, and physical addresses. "Framework is investigating whether or not this included Framework for Business customers as well." TechCrunch reports: Framework's spokesperson Eric Schumacher told TechCrunch that the breach affected "all customers," but declined to specify a specific number. Framework computers are relatively niche products, but some estimates say the company sold hundreds of thousands of devices. Metabase disclosed its own breach in a blog post on its official website, where it said that it was hacked by someone using an unknown security flaw, a so-called zero-day. The company said the hackers exploited the bug to give them the ability to access customers' databases stored on Metabase's cloud servers. In its email to customers, Framework also included the email Metabase sent to the company, which says hackers accessed Framework's cloud instance. The computer maker said it investigated the incident and found that hackers had stolen its customers' personal data, but did not include their payment information.

Read more of this story at Slashdot.

Suno says it will begin watermarking and fingerprinting AI-generated songs, tighten download policies to curb mass distribution, and explicitly prohibit deceptive audio and unauthorized voice or likeness cloning. The changes come as the company faces mounting copyright lawsuits, a recent adverse ruling in Germany, and scrutiny over a past data breach that revealed training-data sources. TechCrunch reports: In a blog post, co-founder and CEO Mikey Shulman shared core principles and said that the platform wants to promote original creation while enabling more people to make music with its AI tools. One of the key points of contention involved users uploading AI-generated songs on other streaming platforms and gaming the system to earn revenue. Suno said that now it will use audio watermarking and fingerprinting to prevent misuse on other streaming platforms. It's not clear if Suno will use an existing system like Google's Synth ID or adopt a new one, and the company did not say when contacted by TechCrunch about this. The startup also signed an agreement with lyrics provider Musixmatch to use its Sentinel system for copyright detection, the blog post said. [...] The company added that it plans to add a new download policy to bar mass distribution on streaming platforms, but declined to provide details on the record. Suno has also changed its community guidelines to explicitly prohibit "deceptive audio presented as real" and "using a real person's voice or likeness without permission" to prevent copycats.

Read more of this story at Slashdot.

Les vieilleries sont de retour avec la crise de la mémoire. Chez NVIDIA, on n'a pas trouvé mieux que de faire revivre une 3e fois la RTX 3060 12 Go, une carte Ampere de la belle époque des cryptomonnaies mais dotée d'un atout irrésistible : ses 12 Go de VRAM qui semblent too much pour une carte dont...

Dans le milieu du refroidissement, Noctua est un peu l’excentrique de la bande. Non contente de proposer des produits peinturlurés en marron et beige reconnaissables entre tous, la marque autrichienne publie aussi des guides dans lesquels elle enjoint ses clients à disposer des ventilateurs partout dans leur logis... [Tout lire]

ByteDance is reportedly training an AI model with roughly 10 trillion parameters as it tries to close the gap with leading frontier systems such as Anthropic's Mythos. The model is still in early pre-training, and its eventual performance will depend on more than scale alone, but the project underscores how aggressively Chinese firms are pushing frontier AI despite limits on access to advanced chips. The Next Web reports: The size is itself the statement. At roughly 10 trillion parameters, the model would be more than three times as large as Moonshot's Kimi K3, which sits among the biggest Chinese models today at about 2.8 trillion. [...] Parameter count is not everything, of course. Bigger models are not automatically better, and the industry has learned that data quality, training technique and efficiency often matter as much as raw scale. Even so, committing the compute to train a model this size is a declaration in its own right, a signal that ByteDance wants to compete at the very top rather than ship a capable also-ran.

Read more of this story at Slashdot.

Intel today released an updated version of their QuickAssist Technology library (QATlib) as the user-space component for making use of QuickAssist accelerators under Linux...

On connaissait les mods physiques à coups de fer à souder pour doubler la quantité de VRAM d'une carte graphique, mais cela implique généralement d'avoir sous la main un nouveau PCB pouvant accueillir deux fois plus de puces mémoire, ces puces mémoire justement, et en ce moment elles ne sont pas don...

This week a final batch of DRM-Misc-Next feature material was submitted for DRM-Next to queue ahead of the Linux 7.3 merge window opening later in the month. Most notable is the TTM memory management code is now being more agressive when allocating below protection limits. This is an improvement coming thanks to Valve's open-source Linux graphics team...

Le programme TV du week-end des 8-9 août tient sa grande affiche jusqu’au bout. Le Tour de France Femmes avec Zwift boucle son édition à Nice, pendant que le MotoGP débarque à Silverstone, en exclusivité sur Canal+. Voici le guide complet des deux journées. Samedi : le Tour de France à Nice Samedi : MotoGP […]

L'article Programme TV du week-end 8-9 août : le guide a été publié en premier sur Bbox-Mag

Following AMD making progress with HDMI Fixed Rate Link (FRL) and other HDMI 2.1 functionality for their open-source Linux kernel graphics driver, Intel is out today with a big set of 44 patches working on HDMI 2.1 FRL support for their kernel graphics driver with Meteor Lake hardware and newer...

President Trump has ordered a 15% tariff on imported products made with polysilicon, a key material for semiconductors and solar panels. The new tariff will take effect on December 4th. The Guardian reports: The executive order signed by the president on Thursday evening said: "The plan of action in this proclamation will, among other things, help ensure the commercial viability of United States production of polysilicon and its derivatives that is necessary to meet United States economic and national security requirements." Polysilicon, an ultra-pure form of silicon, is an important ingredient in making the semiconductors vital for AI processing power, datacentres and solar power generation. US solar factories have long accused Chinese rivals of dumping cheaper panels on the market, which they say have been enabled by excessive government subsidies and by moving manufacturing to other countries to dodge US tariffs. [...] The US has two main polysilicon factories, including Hemlock Semiconductor, which operates a plant in Michigan and is a joint venture between the US tech company Corning and Japan's Shin-Etsu Handotai. The Munich-based Wacker Chemie runs a factory in Tennessee. Trump said in the order that he had accepted recommendations by the commerce secretary, Howard Lutnick, to set minimum import prices of $21 a kilogram for polysilicon, $100 a kilogram for polysilicon ingots and wafers, $0.22 a watt for solar cells, and $0.38 a watt for solar modules or panels. The order also allows the commerce department to create an incentive program for companies that invest in factories to produce polysilicon or derivative products.

Read more of this story at Slashdot.

La bataille de l'enshittification

Le magazine états-unien a commencé à mettre en place des versions différentes des pages de son site, selon qu’il soit consulté par un navigateur ou par un chatbot IA. L’éditeur peut ainsi optimiser ses articles pour qu’ils soient repris par Claude ou ChatGPT en y insérant de la publicité spécifiquement conçue pour être recrachée par les chatbots.

Si les chatbots IA viennent chercher leurs réponses sur les sites des médias, pourquoi ceux-ci n’en profiteraient pas ? C’est ce que s’est dit le magazine américain Time en mettant en place un système servant des pages différentes selon que son site web est consulté par un navigateur habituel ou par un crawler d’un chatbot IA.

L’éditeur a converti toutes les pages de son site web en Markdown, format qui est plus facilement analysable par les agents IA, expliquait Digiday début juillet. De fait, le langage de balisage léger inventé par John Gruber avec l’aide d’Aaron Swartz est idéal pour ces outils : il est explicite et minimise le nombre de tokens utilisés pour analyser le contenu. Si on veut que son contenu soit repris par ChatGPT et Claude, il est donc plutôt malin de le leur servir en Markdown.

La pratique de la redirection sélective pour des buts problématiques n’est pas nouvelle. Par exemple, nous avions déjà montré qu’elle était utilisée pour cacher de la désinformation ou pour adapter des liens sponsorisés en fonction d’attentes parfois nauséabondes d’internautes.

Insérer de la pub spécifiquement rédigée pour les chatbots IA

Ici, l’éditeur a vite compris qu’il pouvait proposer aux marques d’autres services en s’appuyant sur cette technique : le magazine a commencé à glisser des publicités dans certains des fichiers Markdown. Cela permet de faire passer discrètement des messages dans ChatGPT et Claude sans qu’ils n’apparaissent dans les pages proposées aux lecteurs de Time. L’ingénieur allemand Vincent Schmalbach a remarqué le tour de passe-passe et l’a décrit sur son blog.

Ainsi, un article comme celui-ci est proposé en HTML pour tous les navigateurs, mais si on l’ouvre avec le user-agent ClaudeBot, on obtient un texte du genre :

Ici, pas de publicité intégrée. Mais lorsqu’on fait la même opération sur une page listant les « meilleures inventions de 2025 », on se retrouve, contrairement à un lecteur lambda qui l’ouvrirait avec un navigateur, avec un texte promouvant la banque en ligne Ally Bank avant les inventions :

Le texte sur Ally est idéalement écrit pour être cité par un chatbot IA à qui l’utilisateur demanderait « Qui est Ally Bank ? ». Remarquons qu’il est surmonté d’une remarque affirmant : « Contenu sponsorisé. Réalisé en partenariat avec Ally. Ally est le sponsor et la source de ce contenu ».

On obtient le même texte, sous une présentation qui diffère un peu si on utilise les user-agent d’OpenAI GPTBot ou OAI-SearchBot. Remarquons que ce n’est pas le cas si on utilise les crawlers de Google Googlebot, Google-Extended ou GoogleOther mais que le Markdown s’affiche pour le quatrième crawler de Google utilisé par Gemini Google-CloudVertexBot.

Une pratique qui risque de se généraliser

Comme le fait remarquer Vincent Schmalbach, on retrouve dans le fichier Markdown servi à ClaudeBot les traces de l’agence de publicité en ligne Mobian, spécialisée dans la pub pour les agents IA. La page d’accueil de son site reprend spécifiquement l’exemple du Time :

Si nous n’avons pas encore d’exemple d’autres éditeurs de sites très visités qui auraient déjà mis en place ce système, il y a de fortes chances que Time soit rapidement suivi par d’autres. Reste à savoir comment vont réagir les entreprises d’IA générative qui vont devoir faire attention à ce que leurs outils ingèrent.

Va-t-on assister à une nouvelle partie du jeu du chat et de la souris pour bloquer publicités et autres contenus différents de ce qui est affiché dans les pages communément lues par les humains ?

La bataille de l'enshittification

Le magazine états-unien a commencé à mettre en place des versions différentes des pages de son site, selon qu’il soit consulté par un navigateur ou par un chatbot IA. L’éditeur peut ainsi optimiser ses articles pour qu’ils soient repris par Claude ou ChatGPT en y insérant de la publicité spécifiquement conçue pour être recrachée par les chatbots.

Si les chatbots IA viennent chercher leurs réponses sur les sites des médias, pourquoi ceux-ci n’en profiteraient pas ? C’est ce que s’est dit le magazine américain Time en mettant en place un système servant des pages différentes selon que son site web est consulté par un navigateur habituel ou par un crawler d’un chatbot IA.

L’éditeur a converti toutes les pages de son site web en Markdown, format qui est plus facilement analysable par les agents IA, expliquait Digiday début juillet. De fait, le langage de balisage léger inventé par John Gruber avec l’aide d’Aaron Swartz est idéal pour ces outils : il est explicite et minimise le nombre de tokens utilisés pour analyser le contenu. Si on veut que son contenu soit repris par ChatGPT et Claude, il est donc plutôt malin de le leur servir en Markdown.

La pratique de la redirection sélective pour des buts problématiques n’est pas nouvelle. Par exemple, nous avions déjà montré qu’elle était utilisée pour cacher de la désinformation ou pour adapter des liens sponsorisés en fonction d’attentes parfois nauséabondes d’internautes.

Insérer de la pub spécifiquement rédigée pour les chatbots IA

Ici, l’éditeur a vite compris qu’il pouvait proposer aux marques d’autres services en s’appuyant sur cette technique : le magazine a commencé à glisser des publicités dans certains des fichiers Markdown. Cela permet de faire passer discrètement des messages dans ChatGPT et Claude sans qu’ils n’apparaissent dans les pages proposées aux lecteurs de Time. L’ingénieur allemand Vincent Schmalbach a remarqué le tour de passe-passe et l’a décrit sur son blog.

Ainsi, un article comme celui-ci est proposé en HTML pour tous les navigateurs, mais si on l’ouvre avec le user-agent ClaudeBot, on obtient un texte du genre :

Ici, pas de publicité intégrée. Mais lorsqu’on fait la même opération sur une page listant les « meilleures inventions de 2025 », on se retrouve, contrairement à un lecteur lambda qui l’ouvrirait avec un navigateur, avec un texte promouvant la banque en ligne Ally Bank avant les inventions :

Le texte sur Ally est idéalement écrit pour être cité par un chatbot IA à qui l’utilisateur demanderait « Qui est Ally Bank ? ». Remarquons qu’il est surmonté d’une remarque affirmant : « Contenu sponsorisé. Réalisé en partenariat avec Ally. Ally est le sponsor et la source de ce contenu ».

On obtient le même texte, sous une présentation qui diffère un peu si on utilise les user-agent d’OpenAI GPTBot ou OAI-SearchBot. Remarquons que ce n’est pas le cas si on utilise les crawlers de Google Googlebot, Google-Extended ou GoogleOther mais que le Markdown s’affiche pour le quatrième crawler de Google utilisé par Gemini Google-CloudVertexBot.

Une pratique qui risque de se généraliser

Comme le fait remarquer Vincent Schmalbach, on retrouve dans le fichier Markdown servi à ClaudeBot les traces de l’agence de publicité en ligne Mobian, spécialisée dans la pub pour les agents IA. La page d’accueil de son site reprend spécifiquement l’exemple du Time :

Si nous n’avons pas encore d’exemple d’autres éditeurs de sites très visités qui auraient déjà mis en place ce système, il y a de fortes chances que Time soit rapidement suivi par d’autres. Reste à savoir comment vont réagir les entreprises d’IA générative qui vont devoir faire attention à ce que leurs outils ingèrent.

Va-t-on assister à une nouvelle partie du jeu du chat et de la souris pour bloquer publicités et autres contenus différents de ce qui est affiché dans les pages communément lues par les humains ?