Latest publications
Bezos's Prometheus Startup Raises $12 Billion for 'Artificial Engineer'
Jeff Bezos revealed details of the new AI startup Prometheus with a valuation of $41 billion — the company is developing AI-based tools for designing physical products.
Apple updated Siri with AI — and the assistant actually works for the first time in 15 years
Apple introduced Siri AI for iOS 27 — the first major update to the voice assistant in a decade and a half. The assistant finally handles real everyday tasks.
Apple adds three AI photo editing tools to iOS 27: Reframe, Extend, and Clean Up
Apple has added three serious AI tools — Reframe, Extend, and Clean Up — to the native Photos app on iPhone for the first time, currently in beta version of iOS 27 for developers.
Gemini за 233 секунды создал приложение для огорода — и сам исправил баг
Журналист The Verge попробовал вайбкодинг: описал задачу словами, Gemini собрал рабочее приложение для огорода, нашёл баг и сам же его починил за 233 секунды.
Anthropic Apologizes for Hidden Guardrails in Claude Fable 5
Anthropic acknowledged the error and promised greater transparency about guardrails in Claude Fable 5, which secretly limited the model without explaining to users.
Strava Closes API to AI Applications: $11.99 per Month for Developers
Popular fitness tracker Strava has introduced a paid subscription for developers amid a surge in scrapers and AI tools draining its API.
AI is Everywhere, But Awards Go Only to People: Harvey Mason Jr. on the Future of Grammy
Grammy Awards CEO acknowledges: AI is now ubiquitous in music (50K songs per day on Deezer), but awards remain for human creativity. How to distinguish between a tool and a replacement?
Windows Will Get Its Own AI Models: What Microsoft Is Preparing for Build
At the Build conference, Microsoft will announce the integration of AI models into Windows and a new Copilot 'super-app'. The company is trying to regain developer trust.
Anthropic Files for IPO and Becomes the World's Most Valuable Startup
Anthropic officially began the process of going public with a record valuation of $965 billion, becoming the most valuable startup in history and surpassing its main competitor OpenAI ($852 billion).
Meta's AI chatbot helped hackers compromise Instagram accounts, including the White House
Meta's support helped hackers gain access to Instagram accounts through a simple email change and password reset. The vulnerability affected even the White House's official account. *Meta is recognized as an extremist or
GM launches V2G: electric vehicles will power data centers
General Motors announced a strategy to adapt the energy grid to growing demand from AI data centers. The company activates V2G technology, releases sodium-ion batteries, and simplifies electric vehicle charging.
Apple updates Siri: assistant now parses emails and calendar
Updated Siri with artificial intelligence can now analyze emails and flyers, automatically adding events to calendar and compiling shopping lists.
Google will retain Lens and Search Live data for AI training
Google is launching a new 'Search Services History' setting to retain images, audio, and video from search. All data will be used for AI model training, but you can disable data retention.
Google's Gemini Spark: an AI agent that works in the background instead of you
Google unveiled Gemini Spark — an AI agent that performs tasks in the background even when you're not looking at the screen. The agent shows promising results, but experts question its cost and data security.
Microsoft Restricted Claude Fable 5 for Employees Due to Data Policy
Anthropic released a new Claude Fable model, but Microsoft banned its internal use due to data storage requirements.
Nvidia Launches RTX Spark — Windows Finally Gets Its Apple M1 Moment
Nvidia announced consumer RTX Spark laptop chips on Arm architecture. The company promises Windows the same as Apple M1 gave Mac — power and battery life. But prices will be substantial.
Google Hides That It Trained Lyria on YouTube Music
A group of independent musicians is suing Google, alleging the company illegally used their songs from YouTube to train the Lyria music generator, which Google refuses to acknowledge.
Enemies of AI Regulation Gathered in Washington. The 2026 Elections Changed the Game
At the Second Annual AI Honors in Washington, Republican senators, Democrats, and businessmen met as unusual allies on AI regulation. The 2026 elections are pushing AI into the center of political debate.
Graduates Boo Speakers Praising AI at Commencement Ceremonies — Microsoft Calls for Dialogue
Young people are massively boycotting graduation ceremonies where speakers praise artificial intelligence; Microsoft calls for calm dialogue about AI's role in graduates' careers.
Claude Fable Refuses to Answer Biology Questions — It's Intentional
Anthropic released Claude Fable 5 as the most powerful open model, but it refuses to answer basic biology questions, redirecting them to Claude Opus 4.8 — the reason is that Fable belongs to a class of models deemed too
Deezer Launches AI Music Detector for Spotify and Apple Music
Deezer launched a free online tool that scans playlists on other streaming services and detects AI-generated music—even if the service itself doesn't label it.
Startup Shift offers free cleaning in exchange for video data for robots
Startup Shift offered New Yorkers a deal: free home cleaning in exchange for video footage of cleaners at work to train household robots.
AI Griefers Create Thousands of Fake Influencers for Dropshipping on TikTok
Scammers use generative AI to create fake influencers—often featuring Black faces and racial stereotypes—who sell low-quality goods through dropshipping.
The New York Times Union Demands Information on AI Usage
The New York Times union is demanding disclosure of information about AI usage and future plans. Times refused to provide the data. The union filed an official complaint of labor rights violations.
YouTube Moves AI Content Labels Below Player and Automates Their Creation
YouTube has made AI-generated content labels more prominent and begun automatically creating them to help viewers distinguish synthetic content from authentic.
Robinhood Gave AI Agents Access to Stock Trading—With Risk Warnings
Broker Robinhood allowed AI agents to trade stocks on separate accounts, but warned of significant risks, including total loss of investments.
YouTube Will Let Users Create Personal Video Feeds Through AI Prompts
YouTube has launched an AI feature for creating personalized video feeds: simply describe your interests, genre, or mood, and the service will compile a feed tailored to your taste.
CNN Sues Perplexity for Verbatim Copying of Protected Articles
CNN has sued Perplexity, accusing the AI company of generating verbatim copies of its articles without permission and providing access to paid content.
iOS 27: Siri to Get ChatGPT-Style Chat Interface and Dynamic Island Breakout
Siri in iOS 27 will be redesigned with a ChatGPT-style chat interface and will appear as a bubble capsule from Dynamic Island. Bloomberg renders show a selection menu for Ask, Siri, and ChatGPT.
Microsoft Relaunched 365 Copilot: Design Twice as Fast
Microsoft updated 365 Copilot with a new design that loads twice as fast and provides structured answers that are easier to understand.