Ultimate PocketIA
IA
OpenAI rachète pour 7 milliards de dollars d’actions de ses salariés avant son entrée en bourse
En finançant elle-même ce rachat plutôt que de faire appel à des investisseurs, OpenAI évite de faire entrer de nouveaux actionnaires dans son capital à quelques mois d’une possible entrée en Bourse. OpenAI a finalisé une opération de rachat portant sur environ 7 milliards de dollars d’actions détenues par des salariés actuels et anciens, menée directement par l’entreprise, sans recours à des investisseurs externes, à une valorisation inchangée de 852 milliards de dollars. L’information,…
Lire l’article complet
IA
The AI takeover of mathematics has begun
OpenAI showed that AI can tackle long-standing problems in mathematics. Experts are excited about the possibilities — and worried about what comes next for their field. If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement. If you buy something from a Verge link, Vox Media may earn a commission. See our ethics statement. Mathematician James Maynard has spent a lot of time this past year…
Lire l’article complet
IA
AI Is Dead. Organoids Are Alive
I’m going to let you in on a secret. Every cell in your body has the potential to get smarter. I don’t mean this metaphorically, or in a “body keeps the score” kind of way. I mean that if lab-coated biologists took a sample of your skin and very carefully manipulated the cells inside it, they could actually make a brain. They do it all the time. Not a brain as complex as the…
Lire l’article complet
IA
AI Is Helping Solve the Intricate Genetic Puzzle of Schizophrenia
One of the greatest challenges in modern genetics is schizophrenia. Unlike diseases caused by a single mutation, this disorder appears to arise from a combination of hundreds of genetic variants, each with small effects on different brain processes. Some influence neural development, while others alter communication between neurons or the organization of brain connections. Together, they contribute to the risk of developing the disease. Understanding this complex genetic architecture requires more than just identifying…
Lire l’article complet
IA
IA Act : Anthropic impose le filigranage sur Claude (et pourquoi cela ne suffira pas)
Sous la pression de l'IA Act européen, Anthropic annonce l'intégration d'un filigrane invisible sur l'ensemble des contenus générés par ses futurs modèles Claude. La réglementation européenne impose désormais son tempo aux géants de l'intelligence artificielle. Pour se conformer au code de bonnes pratiques sur la transparence des contenus générés par l'IA (Article 50 de l'IA Act), Anthropic vient de confirmer que tous les nouveaux modèles Claude intègreront nativement des outils de traçabilité. Cette politique…
Lire l’article complet
IA
Meta revient aux poids ouverts avec Muse Glimmer, un modèle IA local
Mark Zuckerberg en pleine montée d'IA. Meta reprend son bâton de pèlerin pour les poids ouverts, après l’avoir perdu suite au flop de Llama 4. Muse Glimmer est un nouveau modèle IA de 30 milliards de paramètres, installable sur un ordinateur personnel (assez costaud tout de même), assorti d’une licence open source permissive. En publiant la première version de Muse Spark au mois d’avril, Meta changeait son fusil d’épaule sur la diffusion de son…
Lire l’article complet
IA
Nvidia et Wall Street créent un fonds de 500 milliards de dollars pour financer l'infrastructure IA
Pour soutenir le coût colossal du déploiement de l'intelligence artificielle, Nvidia s'associe à six géants de la finance mondiale. L'objectif de cette alliance à 500 milliards de dollars : transformer l'accès à la puissance de calcul en une classe d'actifs à part entière, tout en diluant le risque financier pour ses clients finaux. L'infrastructure nécessaire pour faire tourner l'économie de l'intelligence artificielle atteint des proportions inédites. Nvidia a annoncé lundi la création d'une plateforme…
Lire l’article complet
IA
OpenAI reportedly completed a $7 billion employee tender offer
OpenAI has bought back $7 billion worth of shares from employees at the privately held frontier AI lab as part of an effort to provide liquidity to its workforce. The deal, reported by Bloomberg, valued OpenAI at $852 billion, the same as its most recent fundraising round in March, which added $122 billion to the company’s war chest. The company also filed confidentially with the Securities and Exchange Commission in June to prepare for…
Lire l’article complet
IA
As AI-led attacks multiply, OpenAI launches a new cyber model
Every day seems to brings fresh news of an AI agent going “rogue.” Whether that’s compromising Hugging Face, hacking a gym website, or creating its own fake profiles to socially engineer an intrusion, AI models are increasingly behaving like bad actors. So, the AI labs that make the models doing the hacking are expanding their cyber protection offerings. This week, OpenAI announced an expansion of Daybreak, its cyber defense service which it launched earlier…
Lire l’article complet
IA
ChatGPT users can now book tables and join restaurant waitlists through Yelp
ChatGPT users can now book restaurant reservations and join waitlists through Yelp without leaving their chats. Here are the details. A few weeks ago, Yelp and ChatGPT struck a content integration partnership that brought Yelp reviews, ratings, photos, and other business information directly into ChatGPT responses. Today, Yelp announced an expansion of that integration, which will let users in the U.S. and Canada book tables and join waitlists through Yelp without leaving the chat.…
Lire l’article complet
IA
With new open models, Meta pitches another reboot of its struggling AI strategy
Meta has announced its intention to focus on open-weight large language models. Additionally, the company announced the release of an open model called Muse Glimmer and a promise to open the weights for Muse Spark 1.2, its more powerful model, in the next few weeks. Alongside these announcements, Meta CEO Mark Zuckerberg published a more than 6,000-word essay outlining the company’s philosophy about AI systems and governance moving forward. The essay aims to differentiate…
Lire l’article complet
IA
Got an ASUS laptop? You might be sitting on a free Google AI Pro or Plus subscription
Affiliate links on Android Authority may earn us a commission. Learn more. If you’re looking to buy a new laptop to take advantage of the latest AI features, you need to be smart. While a large number of new laptops are advertised as “supporting” the latest AI features, there’s often a catch: Getting the most out of those AI features requires you to opt into a paid subscription. Thankfully, a new ASUS promo can…
Lire l’article complet
IA
ChatGPT can now help secure your spot at a restaurant
Affiliate links on Android Authority may earn us a commission. Learn more. ChatGPT can be used for a variety of purposes, from answering questions to helping you vibe code. Recently, OpenAI and Yelp entered into a partnership that allows the chatbot to surface reviews, photos, ratings, and other business details to help you discover local recommendations. Now the two are deepening their partnership to bring a new capability to the chatbot. Yelp has announced…
Lire l’article complet
IA
Mark Zuckerberg’s AI manifesto is exactly why people don’t like AI
On Monday, Mark Zuckerberg published a 6,500-word manifesto about personal AI, largely about the possibilities for the “personal superintelligence” systems Meta AI is building. The ideas in the post aren’t totally new. A version of the essay ran in the Wall Street Journal two weeks ago, and he’s talked about them on Meta earnings calls before. But this is probably the most detailed version he’s shared. I’ve seen it described as an anti-doomer essay,…
Lire l’article complet
IA
Meta’s latest AI model wants to live on your PC
TL;DR Meta’s Muse Glimmer is a 30-billion-parameter AI model designed to run locally on a Mac or PC. It focuses on agentic AI, with the ability to code, use tools, handle multi-step tasks, and recover from errors. The model is available now on Hugging Face, with support for popular local AI tools coming soon. Most modern, capable AI models rely on cloud computing to give you fast and reliable responses — your prompt and…
Lire l’article complet
IA
AI professors are negotiating the new realities of academic research
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Last week, I headed 30 miles south of San Francisco to a hotel in Mountain View, California, to join some of the most accomplished, and some of the most promising, AI researchers in the world. I was hosting roundtable interviews and speaking at a media training for a convening of…
Lire l’article complet
IA
Latest iOS 27 Beta Expands Siri AI Voice Customization
With the fifth beta of iOS 27, Apple has expanded the voice customization options for Siri AI. Pace and Expressivity can now be adjusted for more of the available voice options instead of just the two American voices. There are seven American voice options and four British voices to choose from, all of which can be customized with different Pace and Expressivity settings. Pace changes talking speed, while expressivity changes emphasis. Siri AI voice…
Lire l’article complet
IA
Meta dévoile Muse Glimmer et Mark Zuckerberg en profite pour dérouler sa « philosophie » de l’IA
Recevez tous les soirs un résumé de l’actu importante avec Le Récap’ L’été est chargé pour Meta Superintelligence Labs (MSL), la division IA de Meta pilotée par Alexandr Wang. Après avoir dégainé Muse Code le 6 août 2026 pour aller chasser sur les terres de Claude Code et Codex, Meta enchaîne ce lundi 10 août 2026, avec la sortie de Muse Glimmer, un modèle open weight taillé pour les agents IA locaux. Mark Zuckerberg…
Lire l’article complet
IA
Four takeaways from Mark Zuckerberg’s massive AI manifesto
The Meta CEO makes a case for unleashing Pandora’s box. The Meta CEO makes a case for unleashing Pandora’s box. Meta CEO Mark Zuckerberg has a lot to say about the idealized future he now envisions for humanity co-existing with artificial intelligence — his latest essay spans more than 6,500 words on the matter. The lengthy manifesto Zuckerberg published on Monday, titled “The Future is for Everyone,” broadly lays out his beliefs about how…
Lire l’article complet
IA
Evolve your marketing with new AI tools
We’re adding AI and agentic experiences across Google Ads and Google Analytics to simplify your workflow and expedite your path to business growth. Your browser does not support the audio element. Great marketing has always required deep measurement expertise to uncover the "why" and strategic agility to execute the "how." That’s why we introduced Ask Advisor 1 — our in-product AI agent across Google’s marketing platforms — to put those answers and actions directly…
Lire l’article complet
IA
Plus d’1 Français sur 3 utilise l’IA pour draguer sur les applications de rencontre
De plus en plus de Français confient la séduction sur les applications de rencontre à une IA. Et la plupart des utilisateurs n’assument pas. L’IA peut constituer un bon outil de productivité lorsqu’elle est bien utilisée, mais de nombreuses personnes y ont aussi recours pour draguer sur les applications de rencontre. “Cet été, plus d’un tiers des Français (37 %) avouent avoir fait appel à une intelligence artificielle pour séduire”, apprend-on d’une enquête conduite…
Lire l’article complet
IA
The Download: AI agents for science, and the “censorship-industrial complex”
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. AI for science needs reasoning, not just data —Eric Schmidt, the former CEO of Google and the cofounder of Schmidt Sciences, and Suhas Mahesh, who leads the AI for science work at the AI Center of Schmidt Sciences In 2024, Google DeepMind scientists shared the Nobel Prize in Chemistry for…
Lire l’article complet
IA
OpenAI suspend une partie du développement d’Astra face à des risques cyber inédits
Alors que les capacités cyber des modèles d’IA progressent rapidement, cette pause marque l’une des premières fois qu’un laboratoire de pointe ralentit volontairement le développement d’un de ses modèles pour des raisons de sécurité offensive. OpenAI a annoncé suspendre une partie du développement et des tests de son futur modèle Astra, jugeant ne pas pouvoir exclure qu’il ait atteint le niveau « critique » de capacités en cybersécurité défini par son propre référentiel de…
Lire l’article complet
IA
Claude Cowork, d'Anthropic, se tourne vers le cloud alors que les données montrent que 90 % des sessions ne sont pas consacrées au codage
Claude Cowork se tourne vers le Web et le mobile, et de nouvelles données montrent que son utilisation s'étend bien au-delà du codage et du développement logiciel. Peut-il remplacer OpenClaw ? Les points clés à retenir sur les évolutions de Cowork Anthropic a annoncé début juillet étendre Claude Cowork, l’assistant « agentic » destiné aux utilisateurs de Claude, au Web et aux appareils mobiles. Après avoir analysé 1,2 million de sessions Cowork, Anthropic partage…
Lire l’article complet