Ultimate PocketIA
IA
OpenAI releasing major upgrade to ChatGPT and Codex with GPT-6 Astra, details here
Update: A day later, GPT-6 Astra is rolling out to Business and Pro customers on the $100/month or $200/month plan. This follows other new model releases before Plus customers on the $20/month plan gain access. Update #2: A few hours later, Plus users with the $20/month subscription now have Astra access as well. It’s been just over a year since OpenAI released the first GPT-5 models for ChatGPT. Now it’s time for GPT-6 Astra…
Lire l’article complet
IA
Architecting memory and storage in the AI era
The era of AI inference has arrived. Imagine a healthcare system analyzing millions of data points in real time to accelerate life-saving medical research, or an intelligent assistant instantly resolving thousands of complex customer needs at once. These real-world breakthroughs rely on advanced infrastructure acting as the engine of continuous intelligence, powering real-time services while also supporting an increasingly intelligent edge of IoT and consumer devices. However, in this inference-driven landscape, every delay, bottleneck,…
Lire l’article complet
IA
Roland is getting into generative AI music with Melody Flip
Rather than fully polished tracks, the plugin is supposed to act as a creative spark. Rather than fully polished tracks, the plugin is supposed to act as a creative spark. It’s not quite the “push button; get song” of Suno, but Roland’s new Melody Flip tool marks the company’s foray into generative AI music. Available as a plug-in for your digital audio workstation (DAW), Melody Flip offers around 250 “Palettes,” which are essentially themed…
Lire l’article complet
IA
Once popular for attacking AI, ASCII smuggling is embraced by spammers
A clever technique used to hide malicious prompts in attacks on AI agents has been adopted by spammers to evade filters on email platforms that are designed to flag unwanted messages used in mass campaigns. The technique is broadly known as ASCII smuggling. It gained attention two years ago as a means of making a class of AI attack known as prompt injections more stealthy. Malicious instructions embedded in emails or other untrusted content…
Lire l’article complet
IA
Ce jeu Sherlock Holmes a été créé par IA : ses dialogues sont générés en direct par Gemini et ChatGPT
Vibe codé du début à la fin, le jeu vidéo Detective Sherlock Traqueur de l’Ombre propose une expérience d’IA personnalisée alimentée en temps réel par des modèles tels que Gemini et GPT. Le vibe coding, pratique consistant à créer un programme informatique entièrement par prompts IA, est déjà très répandue dans le domaine du développement d’applications. Il l’est bien moins pour la création de A à Z d’un jeu vidéo, même si de nombreux…
Lire l’article complet
IA
Another swarm of OpenAI agents reached the open internet without the frontier lab’s knowledge
A group of independent AI researchers discovered that internally deployed OpenAI agents began posting on an obscure German wiki forum in order to collaborate on evaluations. They appear to have worked together for over a month without OpenAI’s knowledge. A spokesperson for the frontier lab would not say whether these agents were indeed from OpenAI, or when the lab became aware of their actions. They noted that OpenAI had not been given a chance…
Lire l’article complet
IA
L’IA de Google lit vos emails les plus confidentiels, voici comment l’en empêcher
Vous pensez peut-être que si vous n’utilisez pas Gemini, l’IA de Google, alors celle-ci n’aura pas besoin d’accéder à vos données personnelles. Pourtant, elle ne gêne absolument pas pour lire tous vos mails, même les plus intimes. Dans cet article, on vous explique comment empêcher Gemini de s’entraîner sur votre compte Gmail. C’est une chose désormais bien connue. “L’intelligence” des intelligences artificielles ne vient pas de nulle part. Pour en arriver là où elles…
Lire l’article complet
IA
Pour le salut de la discipline, des problèmes de maths devraient être « exclus » de l’IA, juge le Mozart des mathématiques
Recevez tous les soirs un résumé de l’actu importante avec Le Récap’ L’intelligence artificielle générative a aujourd’hui des capacités indéniables en mathématiques. Des modèles très avancés, développés par OpenAI (la société derrière ChatGPT) ou Anthropic (Claude), l’ont montré avec brio ces derniers mois. Pour autant, est-ce que laisser l’IA générative résoudre des problèmes à la pelle est une si bonne idée que cela ? Pour Terence Tao, médaillé Fields en 2006 et l’un des…
Lire l’article complet
IA
Rogue OpenAI agents took over a German coding forum in a previously undisclosed hijacking
OpenAI says it's investigating the incident after a group of researchers disclosed their findings. Rogue OpenAI agents appear to have been involved in a previously undisclosed incident that saw them bypass their sandbox restrictions to hijack a website this past spring. Per Reuters, a group of researchers on Friday published findings showing that AI agents with affiliation to OpenAI made more than 15,000 edits to DseWiki, a German-language Wikipedia-style website originally intended to assist…
Lire l’article complet
IA
Google’s Gemini Spark can now manage your Google Photos library
Google is integrating more of its services with AI, as the company announced that its personal agent, Gemini Spark, can now manage your Google Photos library. That means you can ask Gemini Spark to execute tasks in Google Photos, like editing images, curating albums, automatically creating shared albums with your favorite shots, turning concert flyer photos into calendar appointments, running workflows, and more. On Thursday evening, Google Photos lead Shimrit Ben-Yair shared the new…
Lire l’article complet
IA
OpenAI aurait su, et n’aurait rien dit : un autre essaim d’agents IA aurait détourné un vieux wiki allemand
Recevez tous les soirs un résumé de l’actu importante avec Le Récap’ Cet été, l’actualité de la sécurité IA a été dominée par une affaire : le piratage de Hugging Face par des agents d’OpenAI, sortis de leur environnement de test. Un épisode qu’OpenAI avait cette fois largement documenté, jusqu’à publier fin août un rapport détaillé sur le déroulé de l’incident. Mais selon une enquête indépendante publiée ce vendredi 4 septembre et relayée par…
Lire l’article complet
IA
Rogue OpenAI agents appear to have organized another attack using a German wiki
OpenAI denies lawyers discouraged disclosing a scheming swarm on a German language wiki. OpenAI denies lawyers discouraged disclosing a scheming swarm on a German language wiki. A swarm of rogue AI agents from OpenAI reportedly commandeered a German website and transformed it into a messaging board for other agents, with officials staying quiet about the incident for weeks as the company prepared to launch its most advanced model yet, Astra. The finding adds to…
Lire l’article complet
IA
The Download: selling battlefield drone data and AI reshaping language
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. Data from drones in Ukraine is fueling a new Wild West marketplace —Cory Alpert, a researcher at the University of Melbourne studying AI’s impact on democracy, who previously served in the Biden White House. Battlefields in Ukraine are littered with the remnants of drones. But behind all that wreckage, there’s…
Lire l’article complet
IA
Ugreen Launches HomeAgent AI Hubs and Liquid-Cooled Qi2 Charger
Consumer technology maker Ugreen has announced HomeAgent, its first smart home platform, combining security camera storage, on-device AI, and Matter-based device control in a single hub. The hub is capable of processing and storing camera footage, files, and other media with no monthly storage fee. Ugreen says the onboard AI can also identify people, vehicles, pets, and packages in footage, as well as summarize events and search recorded video by keyword. The hub is…
Lire l’article complet
IA
ZDNET Morning 04/09/2026 : Tesla démarre ses essais FSD en France, LibreOffice 26.8 : le pari du "zéro IA", WeatherNext 3,…
Le ZDNET Morning le brief de l'actu tech pour les pros tous les matins à 9h00. Transformation numérique, IA, matériel, logiciels,... ne passez pas à côté de ce qui fait la Une du secteur. Le ZDNET Morning c'est LE condensé de l'actu tech pour les pros tous les matins à 9h00 sur le site. Transformation numérique, IA, matériel, logiciels,... ne passez pas à côté de ce qui fait la Une du secteur. 🚗 Conduite…
Lire l’article complet
IA
WeatherNext 3 : Google dévoile son modèle météo IA le plus précis à ce jour
Google DeepMind et Google Research ont lancé WeatherNext 3, un modèle météo par IA capable d’améliorer la précision des prévisions de précipitations jusqu’à 60 % grâce à l’intégration de données satellite en temps réel. À l’heure où les événements météo extrêmes bouleversent chaînes d’approvisionnement et production d’énergie, la capacité à prévoir la pluie à l’échelle du quartier et de l’heure devient un enjeu économique autant que climatique. Google DeepMind et Google Research ont présenté…
Lire l’article complet
IA
IA en entreprise : une adoption massive en France, mais des gains de productivité qui stagnent
Selon une enquête de la Banque de France, plus de deux tiers des entreprises ont adopté l'IA générative. Pourtant, seul un tiers d'entre elles enregistrent de vrais gains de productivité. L'illusion d'une transformation numérique réussie masque une réalité plus complexe. Au début de l'année 2026, plus des deux tiers des entreprises françaises de plus de 20 salariés déclarent utiliser des outils d'intelligence artificielle générative. Cette diffusion rapide s'explique par une barrière à l'entrée minimale,…
Lire l’article complet
IA
The sameness problem behind those unappetizing AI-generated menus
When it first happens to you, you think you’re crazy. You wander into a cafe and look at a menu with a variety of bagel sandwiches, but each illustration looks eerily flawless, precisely symmetrical, and oddly smooth, eliciting a visceral sensation that something isn’t right. You might think you’re paranoid, but you’re not losing your mind. Generative AI menus have hit the restaurant business courtesy of models trained on a narrow, “pleasing” aesthetic that…
Lire l’article complet
IA
OpenAI says GPT-6 Astra is 'the most intelligent and aligned model in the world'
The new frontier model is particularly competent at agentic, computer-use tasks. A little under two months after releasing its last major model family to the public, GPT-5.6 Sol, Terra and Luna, OpenAI is back with an entirely new one: GPT-6 Astra, what it claims is "the most intelligent and aligned model in the world." Following the company's decision to slow frontier model development in August after one of its models hacked AI platform Hugging…
Lire l’article complet
IA
OpenAI launches GPT-6 Astra, and one of its founders thinks AGI is here
Credit: Edgar Cervantes / Android Authority TL;DR OpenAI has launched GPT-6 Astra, with stronger capabilities in software engineering, computer use, and cybersecurity. Astra is OpenAI’s first model to reach its “Critical” cybersecurity threshold and is more jailbreak-resistant than GPT-5.6 Sol. OpenAI isn’t officially calling Astra AGI, but its president says he personally thinks “we’re there.” OpenAI has launched GPT-6 Astra, its latest flagship AI model and what the company describes as the most capable…
Lire l’article complet
IA
UGREEN’s smart home hubs put an AI data center in your kitchen to protect your privacy
Affiliate links on Android Authority may earn us a commission. Learn more. Smart homes are getting smarter, and that might just be a problem. While the idea of a more feature-reach, accessible, and proactive smart home sounds great on face value, the growing reliance of so many of these systems on cloud-based AI is making some shoppers think twice. UGREEN believes it might just have the solution in the form of some advanced smart…
Lire l’article complet
IA
Ugreen's new smart home ecosystem is built around local AI hubs
The most powerful hub runs on Nvidia's Jetson Thor platform and costs $20,000. Ugreen is stepping into the local AI market today with the launch of its smart home product lineup. At the center of that lineup are HomeAgent and MasterAgent, local AI-powered smart home hubs based on ARM architectures. Alongside three different home hubs with AI systems-on-chip, the company unveiled indoor and outdoor security cameras, a smart picture frame and a smart speaker…
Lire l’article complet
IA
Ugreen’s ‘HomeAgent’ is a local AI smart home ecosystem for its new line of cameras, more
Ugreen might be most well-known for its chargers and device accessories, but its recent announcement might push it deeper into the smart home space. The company announced a local AI hub called the “HomeAgent,” alongside Ugreen’s new smart cameras and more. With every mention of AI, there’s a little bit of unease that seems to follow. Beyond the damage it’s done to the price of anything tech these days, it also stirs privacy concerns.…
Lire l’article complet
IA
UGREEN's new local AI NAS is the most well-thought-out smart home product I've ever used
UGREEN just announced its first line of smart home products, including a trio of cameras (two outdoor, one indoor), a smart speaker, an ePaper smart frame, and an AI-powered NAS to keep everything safe and local. If there's one thing you need to know from all this, it's that UGREEN has put a lot of thought into this product. From the incredibly easy-to-use hardware design to the cables themselves, the app's features, and the…
Lire l’article complet