Ultimate PocketIA
IA
ZDNET Morning 20/10/2026 : Union européenne face à Starlink, Nvidia scelle un pacte avec Berlin, DeepSeek bridé,…
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. 🛰️ Orange,…
Lire l’article complet
IA
Mamdani says feds have abdicated on AI, urges New Yorkers to push back
Two of New York’s sweethearts, Mayor Zohran Mamdani and activist and “Broad City” star Ilana Glazer, spoke about AI at Webster Hall in Manhattan on Monday night, with Glazer posing questions to the mayor about the future of AI in society. Mamdani encouraged New Yorkers to resist the overwhelming feeling that AI is inevitably coming for their jobs, their kids’ minds, and their dignity, and to take back control instead. His message came at…
Lire l’article complet
IA
Nscale and Anthropic are reportedly in talks for yet another $45B compute deal
A little over a month ago, Anthropic and AI infrastructure provider Nscale announced a $45 billion compute deal. Now the duo may be about to do it again. Bloomberg reports that the two companies are said to be on the verge of announcing yet another $45 billion compute deal, citing people familiar with the matter. Those sources cautioned that talks are still underway and that the details could change. If the deal goes through,…
Lire l’article complet
IA
Embedded AI evaluators won’t be an easy solution, Hugging Face’s policy chief says
AI labs have staked their reputations on model benchmarking — the supposedly neutral and accurate system of evaluations that are designed to measure whether one AI product is better than another. However, the methods behind these evaluations are complicated and still maturing, said Irene Solaiman, Hugging Face’s chief policy officer, during a discussion at TechCrunch Disrupt last week. Testing needs to come together into a more consistent science if the industry is to rely…
Lire l’article complet
IA
Apple Buys AI Audio Editing App Samplab
Apple acquired AI audio editing app Samplab in June, according to a notice shared today by the European Commission. Samplab offered a desktop app and plugins for converting audio to MIDI, detecting and editing chords, stem splitting for isolating different parts of a song, and editing notes in polyphonic audio without affecting the timbre of the original. Apple could fold the company's tools into Logic Pro or GarageBand. Logic Pro already has Stem Splitter…
Lire l’article complet
IA
Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’
Microsoft on Monday introduced new security features designed to protect enterprises from the growing threat of AI agents. The launch comes as Microsoft continues to warn enterprises about the dangers of AI and agents. Perhaps the most notable addition is Guardian Agents, a new feature intended to monitor the AI agents a company uses. Guardian Agents is part of Project Perception, a product introduced in June. Project Perception was Microsoft’s answer to powerful hacking…
Lire l’article complet
IA
CoreWeave’s debt is becoming the whole AI industry’s problem
CoreWeave had a bad week last week. Its bonds slipped into distressed territory by a common definition — when investors demand 1,000 basis points more in yield than they’d get from government bonds. Last Tuesady Chamath “SPAC King” Palihapitiya noticed the distressed debt, and also asked whether the AI buildout was a private credit bubble. Its shares are down more than 20 percent since Tuesday’s close, which is kind of a lot! There even…
Lire l’article complet
IA
Google is giving you more creative control over Gemini Notebook’s video overviews and slides
Credit: Karandeep Singh / Android Authority TL;DR Gemini Notebook is giving users more creative control over Slide Decks and Video Overviews through a feature called “Co-creation.” The platform will now walk the user through the creation flow, allowing them to steer the format and visual style before the final output. Co-creation capabilities will roll out to all users over the upcoming week. Gemini Notebook is a helpful tool for digesting large amounts of information.…
Lire l’article complet
IA
Accused OpenAI spy had folder called “Saudi Gov Docs,” FBI says
On Monday, a court granted the US government’s request to deny release of a former OpenAI engineer accused of bugging a top-secret conference room to steal the frontier lab’s secrets. In a court filing Sunday, US attorney Craig Missakian argued that Abdulrahman Alfozan must continue to be detained while the US makes its case that Alfozan brazenly spied on OpenAI, either for his own personal business gain or to benefit the Kingdom of Saudi…
Lire l’article complet
IA
Apple strikes deal with AI startup that lets musicians rework recorded music
The EU’s List of Acquisitions website was updated today with details of a deal struck in June between Apple and Samplab, a Swiss startup that described itself as offering “music production for the age of AI.” Here are the details. Last week, the List of Acquisitions section of the EU’s DMA website was updated with details on Apple’s move to acqui-hire Huxe, a startup founded by former NotebookLM engineers that specialized in generating personalized…
Lire l’article complet
IA
Making the internet better for AI agents might make it worse for everyone else
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. Until recently, you could divide internet users into just two categories: bots and people. Websites have long allowed bots to perform automated tasks, like capturing the content of web pages, but rules govern what they can access and how they behave. The web, after all, is designed primarily for people,…
Lire l’article complet
IA
“We were looking at it wrong”: The explosive claim that’s upended the AI world
“I think I have good news about the nature of neural networks.” With those simple words of optimism, Nick Cammarata, a researcher at the AI startup Goodfire, kicked off a 26,000-word blog post that has knocked the AI world on its head, sparking wild speculation as well as cautious (and not-so-cautious) celebration. Cammarata’s claim is that LLMs might in fact be far easier to understand—and thus to monitor and control—than people previously thought. If…
Lire l’article complet
IA
☕️ Les États-Unis bannissent l’IA chinoise DeepSeek dans l’administration civile fédérale
Le département états-unien de la Sécurité intérieure (DHS) ordonne le retrait des produits DeepSeek et services tiers utilisant le modèle d’IA chinois des systèmes informatiques des agences civiles fédérales, rapporte l’agence Reuters. Le DHS leur interdit également d’acquérir des produits DeepSeek, ainsi que des services tiers développés à l’aide des outils de l’entreprise chinoise. L’AFP précise que le DHS, qui a autorité sur les services civils du gouvernement (mais pas sur l’armée ni le…
Lire l’article complet
IA
Disney found a way to buy Character.AI without buying its lawsuits
This week, Character.AI declared bankruptcy. While some saw the fall of a chatbot platform as a harbinger for the industry, court filings indicate that the decision was actually a key part of a deal that has let Disney hire most of Character.AI’s staff—including its now former CEO Karandeep Anand—potentially acquire most of Character.AI’s technology, and avoid the beleaguered startup’s many legal liabilities and obligations. In the past few years, Character.AI has faced multiple private…
Lire l’article complet
IA
JB Pritzker thinks AI could help us live to 100 — or threaten us like nukes
Today, I’m talking with Illinois Governor JB Pritzker, who has a lot of thoughts about AI, tech, and privacy. If you’ve been listening to the show, you know that one of my running theories is that those issues are driving a big and maybe kind of unexpected political realignment — and with the midterms coming up next month, and early voting already underway in a lot of places, the stakes are pretty high right…
Lire l’article complet
IA
Apple is right about the wild west of AI, and the growing backlash
Apple marketing chief Greg Joswiak didn’t hold back when he described the most ambitious figures in the AI race as the “lunatic fringe” of Silicon Valley. The piece suggested that the iPhone maker is hoping to position itself as being on the right side of a growing AI backlash. Given the pace of development, trying to predict the future of AI is a risky business, but I do think Apple is absolutely right about…
Lire l’article complet
IA
Quatre IA de Google ont fini Minecraft en 27 heures sans intervention humaine
Selon HowLongToBeat, les joueurs humains mettent en moyenne 45 heures pour finir l’histoire principale de Minecraft. Quatre agents IA de Google l’ont fait en 27 heures et 14 minutes, sans qu’un humain intervienne une seule fois. Trois chercheurs de Google DeepMind, Maxence Faldor, Satinder Singh et Roberta Raileanu, racontent la partie dans une étude mise en ligne le 15 octobre sur arXiv. Chaque agent tourne avec Gemini 4 Argon, le modèle de Google annoncé…
Lire l’article complet
IA
Google support pages start defaulting to AI chatbots – it might be for the best
Google is starting to move its support pages from detailed articles to AI summaries with a chatbot, with an added set of steps to access the full details of the any given page. The Google Help portal is full of detailed articles that provide step-by-step processes for using a feature, documentation on where features are available or restrictions on them, as well as troubleshooting steps if something goes awry. It’s a massively useful resource,…
Lire l’article complet
IA
Vous disposez d'un compte professionnel Meta ? Selon ces juristes, cela revient tout simplement à accepter le « scraping » par l'IA
Et il n'existe aucune possibilité de se désinscrire. Les points clés à retenir sur le scraping de vos données pros par Meta Les nouvelles conditions commerciales de Meta autorisent ses robots d’indexation, ses bots et ses agents à explorer les sites web « associés » des comptes professionnels Meta, y compris ceux des éditeurs d’actualités, selon un rapport du Nieman Lab. Ces conditions s’appliquent aux comptes qui utilisent le tableau de bord de Meta…
Lire l’article complet
IA
The Download: preserving the Arctic permafrost, and playful AI agents
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. Inside the quest to combat Arctic fires and save the planet’s frozen carbon —James Temple In July 2021, a wildfire burned through a forest on the University of Alaska, Fairbanks campus, stripping away the moss and organic-rich soil insulating the permafrost below. Five years later, I followed researchers into the…
Lire l’article complet
IA
L’outil anti-IA de Deezer distingué par TIME parmi les meilleures inventions 2026
Illustration éditoriale originale générée : scène musicale fictive inspirée du thème «L’outil de détection de la musique générée par IA de Deezer figure dans la liste des meilleures inventions de 2026 publiée par TIME ». Elle ne représente aucun véritable artiste ni événement. Deezer annonce, dans un communiqué publié le 15 octobre 2026 sur sa newsroom, que son outil de détection de la musique générée par intelligence artificielle figure dans la liste des meilleures…
Lire l’article complet
IA
Anthropic propose au Congrès américain un projet de loi pour bloquer les IA jugées dangereuses
Anthropic aurait soumis à des élus américains un projet de loi permettant de bloquer un modèle d’IA en cas de risque imminent de plus d’un milliard de dollars de dégâts ou de 50 morts ou blessés graves. Un laboratoire d’IA peut-il écrire la loi qui l’encadre ? Anthropic a présenté à des parlementaires américains un projet de texte qui permettrait de bloquer un modèle jugé dangereux, y compris parmi les plus puissants du marché.…
Lire l’article complet
IA
Edge AI for Smart Cities: From Insight to Cognitive Action
Smart city deployments have historically relied on cameras, sensors, and dashboards that capture data without delivering timely insight. Edge AI changes this by letting inference happen close to where data is generated. This supports faster decisions, stronger data sovereignty, and lower operating costs than cloud-only designs. This guide traces the progression from automation and perception to prediction, agentic coordination, and physical AI, where urban systems continuously perceive conditions, reason about them, and act within…
Lire l’article complet
IA
Google Gemini 4 agents just beat Minecraft with no human intervention
Agents split up tasks, warned about threats and even recovered items from dead teammates. Google's Deepmind "AlphaGo" AI is famous for winning at Go, but the company's latest Gemini AI just mastered a far more complex game. Four Gemini 4 Argon agents beat Minecraft's final "Ender Dragon" foe with no human intervention by working individually while sharing tips, threats and mistakes, Google researchers revealed in a new paper. The experiment shows how general agentic…
Lire l’article complet