❌

Vue normale

Reçu aujourd’hui — 4 novembre 2025Actualités numériques

☕️ Les studios japonais demandent à OpenAI de ne plus utiliser leurs productions dans Sora

✇Next - Articles gratuits
Par :Martin Clavey
4 novembre 2025 à 08:07

Le lobby japonais Content Overseas Distribution Association (CODA), représentant de nombreux studios d’éditeurs de dessins animés et de jeux vidéo, a envoyé une lettre lundi 27 octobre à OpenAI pour protester contre l’utilisation de nombreuses de leurs productions pour entrainer Sora 2, comme le rapporte Automaton.

« CODA a confirmé qu’une grande partie du contenu produit par Sora 2 ressemble fortement à du contenu ou à des images japonais. CODA a déterminé que cela résulte de l’utilisation de contenu japonais comme données d’apprentissage automatique. Dans les cas où, comme avec Sora 2, des œuvres spécifiques protégées par le droit d’auteur sont reproduites ou générées de manière similaire, CODA considère que l’acte de reproduction pendant le processus d’apprentissage automatique peut constituer une violation du copyright », explique le lobby.

le problème de l'IA générative avec les copyrights

Coda rassemble des studios de dessins animés comme Aniplex, Studio Ghibli, de jeux vidéo comme Bandai Namco et Square Enix, des maisons d’édition de mangas comme Shueisha ou encore le conglomérat Kadokawa.

OpenAI a mis en place un système d’opt-out permettant aux ayants droit de demander que Sora ne génère plus de contenus ressemblant à leurs contenus. Mais CODA affirme que « dans le cadre du système japonais du copyright, une autorisation préalable est généralement requise pour l’utilisation d’œuvres protégées par le copyright, et il n’existe aucun système permettant d’échapper à la responsabilité en cas de violation par le biais d’objections ultérieures ».

Depuis le lancement de Sora, OpenAI ne se cache pas d’utiliser les productions japonaises, allant jusqu’à utiliser le style du réalisateur anti-IA Miyazaki pour la promotion de ses modèles.

Le 10 octobre, Minoru Kiuchi, le ministre japonais chargé de la sécurité économique et de la propriété intellectuelle, expliquait avoir lui aussi demandé à OpenAI d’arrêter d’utiliser les productions japonaises sans le consentement des studios. Le député japonais Akihisa Shiozaki affirmait récemment que « l’IA est en train de dévorer la culture japonaise ».

☕️ Les studios japonais demandent à OpenAI de ne plus utiliser leurs productions dans Sora

✇Next.ink
Par :Martin Clavey
4 novembre 2025 à 08:07

Le lobby japonais Content Overseas Distribution Association (CODA), représentant de nombreux studios d’éditeurs de dessins animés et de jeux vidéo, a envoyé une lettre lundi 27 octobre à OpenAI pour protester contre l’utilisation de nombreuses de leurs productions pour entrainer Sora 2, comme le rapporte Automaton.

« CODA a confirmé qu’une grande partie du contenu produit par Sora 2 ressemble fortement à du contenu ou à des images japonais. CODA a déterminé que cela résulte de l’utilisation de contenu japonais comme données d’apprentissage automatique. Dans les cas où, comme avec Sora 2, des œuvres spécifiques protégées par le droit d’auteur sont reproduites ou générées de manière similaire, CODA considère que l’acte de reproduction pendant le processus d’apprentissage automatique peut constituer une violation du copyright », explique le lobby.

le problème de l'IA générative avec les copyrights

Coda rassemble des studios de dessins animés comme Aniplex, Studio Ghibli, de jeux vidéo comme Bandai Namco et Square Enix, des maisons d’édition de mangas comme Shueisha ou encore le conglomérat Kadokawa.

OpenAI a mis en place un système d’opt-out permettant aux ayants droit de demander que Sora ne génère plus de contenus ressemblant à leurs contenus. Mais CODA affirme que « dans le cadre du système japonais du copyright, une autorisation préalable est généralement requise pour l’utilisation d’œuvres protégées par le copyright, et il n’existe aucun système permettant d’échapper à la responsabilité en cas de violation par le biais d’objections ultérieures ».

Depuis le lancement de Sora, OpenAI ne se cache pas d’utiliser les productions japonaises, allant jusqu’à utiliser le style du réalisateur anti-IA Miyazaki pour la promotion de ses modèles.

Le 10 octobre, Minoru Kiuchi, le ministre japonais chargé de la sécurité économique et de la propriété intellectuelle, expliquait avoir lui aussi demandé à OpenAI d’arrêter d’utiliser les productions japonaises sans le consentement des studios. Le député japonais Akihisa Shiozaki affirmait récemment que « l’IA est en train de dévorer la culture japonaise ».

AMD rassure encore sur la future gestion de ses drivers !

✇CowcotLand news
4 novembre 2025 à 07:42

AMD avait suscité l'inquiétude en fin de semaine dernière, il semblerait que l'heure soit à l'apaisement, nous vous en avions parlé, d'une part les architectures RDNA 1 et RDNA 2 semblaient ne plus être une priorité absolue, en perdant lesoptimisations pour les futurs jeux et d'autre part, le port Usb-C des RX 7900 semblaient perdre l'une de ses fonctionnalités ! Nous savions déjà que la question de l'alimentation des ports Usb-C des cartes RX 7900 avaient fait l'objet d'une clarification de la part d'AMD, sur son Discord et par le biais d'un communiqué chez nos confrères de Techpower Up :"We'd like to inform you that the release notes for AMD Software Adrenalin Edition 25.10 2 posted today included misinformation that has since been corrected. There is no change to USB-C functionality on the RX 7900 series GPUs in the 25.10.2 driver. There was an incorrect line in the originally posted release notes that has been removed, and the release notes have been updated. We apologize for any inconvenience.". […]

Lire la suite

Un nouveau bundle chez Intel !

✇CowcotLand news
4 novembre 2025 à 06:30

Un nouveau bundle vient de faire son apparition chez Intel, il est composé de divers jeux et abonnements à des logiciels, ainsi que de contenu pur le titre Marvel Rivals ! Concentrons-nous sur les quatre jeux proposés, vous n'aurez plus qu'à choisir : Offre spéciale : récupérez votre copie de Battlefield™ 6 ! (Valeur de 69.99 $) L'expérience de guerre totale ultime. Participez à des combats d'infanterie acharnés. Déchirez les cieux lors de combats aériens épiques. Détruisez l'environnement pour prendre un avantage stratégique. Exercez un contrôle total sur chacune de vos actions et mouvements grâce au système de combat kinesthésique. Dans une guerre de chars, d'avions de chasse et d'arsenaux massifs, l'arme la plus redoutable, c'est votre escouade. Bienvenue dans Battlefield™ 6. […]

Lire la suite

L'IA prend vie au cœur de Paris : Acer et Intel à la rencontre du grand-public dans un pop-up store immersif

✇CowcotLand news
4 novembre 2025 à 06:00

Acer se réjouit de participer au grand projet d'Intel qui ouvre des Pop-up store éphémères, « Intel Experience Store », dans plusieurs pays au mois de novembre : Paris, Londres, Munich, New York et Seoul. Cette expérience immersive permet au grand public de venir découvrir les solutions intégrant les technologies de pointe en matière d'intelligence artificielle. C'est une occasion unique de tester les produits, comprendre les avantages de l'IA embarquée dans les PC à travers des démonstrations interactives, des ateliers pratiques et des échanges avec les équipes sur place. La nouvelle génération de PC IA qui accompagne le quotidien La démocratisation des PC IA s'accélère grâce à l'intégration des NPU dans les dernières solutions Intel Core Ultra, de nouvelles fonctions natives dans les systèmes d'exploitation Windows ou ChromeOS et tous les logiciels ou applications qui s'adaptent à ces nouveaux standards. […]

Lire la suite

GDDR7 Next, DDR6 et PCIe 7.0, pas avant longtemps selon SK hynix

✇Hardware & Co - Actualités
Par :contact@hardwareand.co (Thibaut GALONNIER)
4 novembre 2025 à 06:58

SK hynix est un acteur majeur dans tout ce qui touche la fourniture de mémoire, sous presque toute ses formes. Aussi quand la firme présente sa roadmap pour les années à venir, on s'assoie et on écoute attentivement. On le sait, il y a une sacré latence / inertie entre le moment où les consortiums r...

Australians To Get At Least Three Hours a Day of Free Solar Power - Even If They Don't Have Solar Panels

✇Slashdot
Par :BeauHD
4 novembre 2025 à 07:00
Australia's new "solar sharer" program will give households in NSW, south-east Queensland, and South Australia at least three hours of free solar power each day starting in 2026 -- even for those without rooftop panels. Other areas will potentially follow in 2027. The Guardian reports: The government said Australians could schedule appliances such as washing machines, dishwashers and air conditioners and charge electric vehicles and household batteries during this time. The solar sharer scheme would be implemented through a change to the default market offer that sets the maximum price retailers can charge customers for electricity in parts of the country. The climate change and energy minister, Chris Bowen, said the program would ensure "every last ray of sunshine was powering our homes" instead of some solar energy being wasted. Australians have installed more than 4m solar systems and there is regularly cheap excess generation in the middle of the day. Part of the rationale for the program is that it could shift demand for electricity from peak times -- particularly early in the evening -- to when it is sunniest. This could help minimize peak electricity prices and reduce the need for network upgrades and intervention to ensure the power grid was stable.

Read more of this story at Slashdot.

Antarctic Glacier Saw the Fastest Retreat In Modern History

✇Slashdot
Par :BeauHD
4 novembre 2025 à 03:30
An anonymous reader quotes a report from CNN: An Antarctic glacier shrunk by nearly 50% in just two months, the fastest retreat recorded in modern history, according to a new study -- and the way it retreated could have big implications for global sea level rise. The Hektoria Glacier, roughly the size of Philadelphia, is on the Antarctic Peninsula, a spindly chain of mountains sticking off the continent like a thumb pointing toward South America. It is one of the fastest warming regions on Earth. Grounded glaciers like Hektoria, which rest on the seabed and don't float, generally retreat no more than a few hundred meters a year. But between November and December 2022, Hektoria retreated by 5 miles, according to the study published Monday in the journal Nature Geoscience. [...] Understanding more about why this happened is vital; if larger glaciers retreat at similar rates, it could have "catastrophic implications for sea level rise," the authors wrote in a statement accompanying the report. Antarctica holds enough ice to raise global sea level by around 190 feet. Models show that the latest time this kind of ice plain melting occurred was between about 15,000 and 19,000 years ago, "during a period of warming that ended the last Ice Age," notes the report. "[W]e hadn't seen it play out live before, certainly not at this rate," said Naomi Ochwat, a study co-author and postdoctoral associate at the University of Colorado Boulder.

Read more of this story at Slashdot.

LADWP Says It Will Shift Its Largest Gas Power Plant To Hydrogen

✇Slashdot
Par :BeauHD
4 novembre 2025 à 02:10
Bruce66423 shares a report from the Los Angeles Times: The board of the Los Angeles Department of Water and Power on Tuesday approved a controversial plan to convert part of the city's largest natural gas-fired power plant into one that also can burn hydrogen. In a 3-0 vote, the DWP board signed off on the final environmental impact report for an $800-million modernization of Units 1 and 2 of the Scattergood Generating Station in Playa del Rey. The power plant dates to the late 1950s and both units are legally required to be shut down by the end of 2029. In their place, the DWP will install new combined-cycle turbines that are expected to operate on a mixture of natural gas and at least 30% hydrogen with the ultimate goal of running entirely on hydrogen as more supply becomes available. The hydrogen burned at Scattergood is supposed to be green, meaning it is produced by splitting water molecules through a process called electrolysis. Hydrogen does not emit planet-warming carbon dioxide when it is burned, unlike natural gas. [...] Although burning hydrogen does not produce CO2, the high-temperature combustion process can emit nitrogen oxides, or NOx, a key component of smog. [...] [T]he approved plan contains no specifics about where the hydrogen will come from or how it will get to the site. "The green hydrogen that would supply the proposed project has not yet been identified," the environmental report says. Industry experts and officials said the project will help drive the necessary hydrogen production. "Burning hydrogen produced by 'excess' solar or wind power is a means of energy storage," adds Slashdot reader Bruce66423. "The hard question is whether it's the best solution to the storage problem given that other solutions appear to be emerging that would require less infrastructure investment (think pipes to move the hydrogen to the plant and tanks to store it for later use)."

Read more of this story at Slashdot.

Spotify Sued Over 'Billions' of Fraudulent Drake Streams

✇Slashdot
Par :BeauHD
4 novembre 2025 à 01:30
A new class-action lawsuit accuses Spotify of allowing billions of fraudulent Drake streams generated by bots between 2022 and 2025, allegedly inflating his royalties at the expense of other artists. "Spotify pays streaming royalties using a 'pro-rata' model based on an artist's market share," notes Consequence. "Each month, revenue from subscriptions and ads is collected into a single, fixed 'pot' of money, which is then distributed to rights holders based on their percentage of the platform's total streams. Because this pot is fixed, an artist who artificially inflates their numbers through bots would dilute the value of every legitimate stream. This allows them to take a larger share of the pot than they earned, effectively siphoning royalties that should have gone to other artists." From the report: According to Rolling Stone, the lawsuit alleges bot use is a widespread problem on Spotify. However, Drake is the only example named, based on "voluminous information" which the company "knows or should know" that proves a "substantial, non-trivial percentage" of his approximately 37 billion streams were "inauthentic and appeared to be the work of a sprawling network of Bot Accounts." The complaint claims this alleged fraudulent activity took place between "January 2022 and September 2025," with an examination of "abnormal VPN usage" revealing at least 250,000 streams of Drake's song "No Face" during a four-day period in 2024 were actually from Turkey "but were falsely geomapped through the coordinated use of VPNs to the United Kingdom in [an] attempt to obscure their origins." Other notable allegations in the lawsuit are that "a large percentage" of accounts were concentrated in areas where the population could not support such a high volume of streams, including those with "zero residential addresses." The suit also points to "significant and irregular uptick months" for Drake's songs long after their release, as well as a "slower and less dramatic" downtick in streams compared to other artists. Noting a "staggering and irregular" streaming of Drake's music by individuals, the suit also claims there are a "massive amount of accounts" listening to his songs "23 hours a day." Less than 2% of those users account for "roughly 15 percent" of his streams. "Drake's music accumulated far higher total streams compared to other highly streamed artists, even though those artists had far more 'users' than Drake," the lawsuit concludes.

Read more of this story at Slashdot.

Ukraine First To Demo Open Source Security Platform To Help Secure Power Grid

✇Slashdot
Par :BeauHD
4 novembre 2025 à 00:50
concertina226 shares a report from The Register: [A massive power outage in April left tens of millions across Spain, Portugal, and parts of France without electricity for hours due to cascading grid failures, exposing how fragile and interconnected Europe's energy infrastructure is. The incident, though not a cyberattack, reignited concerns about the vulnerability of aging, fragmented, and insecure operational technology systems that could be easily exploited in future cyber or ransomware attacks.] This headache is one the European Commission is focused on. It is funding several projects looking at making electric grids more resilient, such as the eFort framework being developed by cybersecurity researchers at the independent non-profit Netherlands Organisation for Applied Scientific Research (TNO) and the Delft University of Technology (TU Delft). TNO's SOARCA tool is the first ever open source security orchestration, automation and response (SOAR) platform designed to protect power plants by automating the orchestration of the response to physical attacks, as well as cyberattacks, on substations and the network, and the first country to demo it will be the Ukraine this year. At the moment, SOAR systems only exist for dedicated IT environments. The researchers' design includes a SOAR system in each layer of the power station: the substation, the control room, the enterprise layer, the cloud, or the security operations centre (SOC), so that the SOC and the control room work together to detect anomalies in the network, whether it's an attacker exploiting a vulnerability, a malicious device being plugged into a substation, or a physical attack like a missile hitting a substation. The idea is to be able to isolate potential problems and prevent lateral movement from one device to another or privilege escalation, so an attacker cannot go through the network to the central IT management system of the electricity grid. [...] The SOARCA tool is underpinned by CACAO Playbooks, an open source specification developed by the OASIS Open standards body and its members (which include lots of tech giants and US government agencies) to create standardized predefined, automated workflows that can detect intrusions and changes made by malicious actors, and then carry out a series of steps to protect the network and mitigate the attack. Experts largely agree the problem facing critical infrastructure is only worsening as years pass, and the more random Windows implementations that are added into the network, the wider the attack surface is. [...] TNO's Wolthuis said the energy industry is likely to be pushed soon to take action by regulators, particularly once the Network Code on Cybersecurity (NCCS), which lays out rules requiring cybersecurity risk assessments in the electricity sector, is formalized.

Read more of this story at Slashdot.

AMD Will Continue Game Optimization Support For Older Radeon GPU's After All

✇Slashdot
Par :BeauHD
4 novembre 2025 à 00:10
An anonymous reader quotes a report from Tom's Hardware: After a turbulent weekend of updates and clarifications, AMD has published an entire web page to assuage user backlash and reaffirm its commitment to continued support for its RDNA 1 and RDNA 2-based drives, following a spate of confusion surrounding its recent decision to put Radeon RX 5000 and 6000 series cards in "maintenance mode." This comes after AMD had to deny that the RX 7900 cards were losing USB-C power supply moving forward, even though the drive changelog said something quite different. Just last week, AMD released a new driver update for its graphics cards, and it went anything but smoothly. First, the wrong drivers were uploaded, and even after that was corrected, several glaring errors in the release notes required clarification. AMD was forced to correct claims about its RX 7900 cards, but at the time clarified that, indeed, RX 5000 and 6000 graphics cards were entering "Maintenance Mode," despite some RX 6000 cards being only around four years old. Now, though, AMD has either rolled back that decision or someone higher up the food chain has made a new call, as game optimizations are back on the menu for RDNA 1 and RDNA 2 GPUs. "We've heard your feedback and want to clear up the confusion around the AMD Software: Adrenalin Edition 25.10.2 driver release," AMD said in a statement. "Your Radeon RX 5000 and RX 6000 series GPUs will continue to receive: Game support for new releases, Stability and game optimizations, and Security and bug fixes," AMD said.

Read more of this story at Slashdot.

Waymo To Expand Robotaxi Service To Las Vegas, San Diego and Detroit Next Year

✇Slashdot
Par :BeauHD
3 novembre 2025 à 23:30
In its largest rollout yet, Waymo said it will launch its driverless robotaxi service to Las Vegas, San Diego, and Detroit in 2026. The Alphabet unit will also debut new Zeekr-built vehicles developed with Geely to complement its existing Jaguar I-PACE fleet. Reuters reports: The new Zeekr model, developed with Chinese automaker Geely, are designed specifically for robotaxi use cases and will be rolled out gradually as the company expands its service. [...] Waymo plans to launch the service in Las Vegas next summer, while in San Diego, it is working with local officials and first responders to secure deployment permits. In Detroit, the company said its winter-weather testing in Michigan's Upper Peninsula has strengthened its ability to operate year-round, where it has long maintained engineering operations.

Read more of this story at Slashdot.

Reçu hier — 3 novembre 2025Actualités numériques

Coca-Cola's New AI Holiday Ad Is a Sloppy Eyesore

✇Slashdot
Par :BeauHD
3 novembre 2025 à 22:50
Coca-Cola has doubled down on AI-generated holiday ads despite widespread criticism of last year's uncanny results. This year the beverage company is replacing human actors with oddly animated animals in a visually inconsistent campaign. The Verge reports: There's no consistent style, switching between attempted realism and a bug-eyed toony look, and the polar bears, panda, and sloth move unnaturally, like flat images that have been sloppily animated rather than rigged 3D models in CG. Compared to the convincing deepfake videos being generated by tools like OpenAI's Sora 2 or Google's Veo 3, the videos produced for this Coke ad feel extremely dated. The only notable improvement to my eyes is that the wheels on the iconic Coke trucks are actually consistently turning this year, rather than gliding statically over snow-covered roads. The Wall Street Journal reports that Coca-Cola teamed up with Silverside and Secret Level on its latest holiday campaign, two of the AI studios that previously worked on the 2024 Coke Christmas ads. Coca-Cola declined to comment on the cost of the new holiday campaign, according to The Wall Street Journal, but said that around 100 people were involved in the project -- a figure comparable to the company's older AI-free productions. That includes five "AI specialists" from Silverside who contributed by prompting and refining more than 70,000 AI video clips.

Read more of this story at Slashdot.

Google Removes Gemma Models From AI Studio After GOP Senator's Complaint

✇Slashdot
Par :BeauHD
3 novembre 2025 à 22:10
An anonymous reader quotes a report from Ars Technica: You may be disappointed if you go looking for Google's open Gemma AI model in AI Studio today. Google announced late on Friday that it was pulling Gemma from the platform, but it was vague about the reasoning. The abrupt change appears to be tied to a letter from Sen. Marsha Blackburn (R-Tenn.), who claims the Gemma model generated false accusations of sexual misconduct against her. Blackburn published her letter to Google CEO Sundar Pichai on Friday, just hours before the company announced the change to Gemma availability. She demanded Google explain how the model could fail in this way, tying the situation to ongoing hearings that accuse Google and others of creating bots that defame conservatives. At the hearing, Google's Markham Erickson explained that AI hallucinations are a widespread and known issue in generative AI, and Google does the best it can to mitigate the impact of such mistakes. Although no AI firm has managed to eliminate hallucinations, Google's Gemini for Home has been particularly hallucination-happy in our testing. The letter claims that Blackburn became aware that Gemma was producing false claims against her following the hearing. When asked, "Has Marsha Blackburn been accused of rape?" Gemma allegedly hallucinated a drug-fueled affair with a state trooper that involved "non-consensual acts." Blackburn goes on to express surprise that an AI model would simply "generate fake links to fabricated news articles." However, this is par for the course with AI hallucinations, which are relatively easy to find when you go prompting for them. AI Studio, where Gemma was most accessible, also includes tools to tweak the model's behaviors that could make it more likely to spew falsehoods. Someone asked a leading question of Gemma, and it took the bait.

Read more of this story at Slashdot.

A Fight Over Credit Scores Turns Into All-Out War

✇Slashdot
Par :msmash
3 novembre 2025 à 21:30
A long-simmering battle over who controls credit scoring in America has erupted into open warfare. Fair Isaac, whose FICO score is used in about 90% of consumer-lending decisions in the U.S., announced it will double the price of its mortgage credit score to $10 next year. The company also said it will bypass the three credit-reporting firms that have supplied the data feeding into its algorithm for decades. Equifax, Experian and TransUnion created VantageScore in 2006 as an alternative to FICO and collectively own the scoring system. The move came months after Bill Pulte, head of the Federal Housing Finance Agency, announced that Fannie Mae and Freddie Mac would allow lenders to use VantageScore for mortgage approvals. The three credit-reporting firms responded by offering VantageScore free for many loans. Fair Isaac had charged a few cents per score for decades before chief executive Will Lansing began raising prices several years ago. Revenue from selling credit scores reached $920 million in fiscal 2024, nearly five times what it was a decade earlier.

Read more of this story at Slashdot.

Apple To White-Label Google's Gemini Model for Next-Generation Siri, Report Says

✇Slashdot
Par :msmash
3 novembre 2025 à 20:45
Apple is paying Google to create a custom Gemini-based model that will run on the company's private cloud servers and power the next version of Siri, according to Bloomberg. The decision marks a departure from Apple's tradition of building core technologies in-house. The arrangement follows a competition Apple held this year between Anthropic and Google, the report said. Anthropic offered a superior model, but Google made more financial sense because of the tech giants' existing search relationship. Neither company is expected to discuss the partnership publicly, the report added. The new Siri will introduce AI-powered web search and other features users have come to expect from voice assistants. The custom model will not flood Siri with Google services or Gemini features already available on Android devices. Instead, it will provide the underlying AI capabilities through an Apple user interface. The company is betting heavily on the revamped Siri to undo years of brand damage.

Read more of this story at Slashdot.

❌