Latest publications
Adobe's Firefly: An Assistant That Explains Well, But Impresses Rarely
Adobe created Firefly AI Assistant — a chat assistant for design that manages applications on behalf of the user. It preserves creative control and explains every step, but the results don't always impress.
Shift startup offers free house cleaning in exchange for robot training videos
Shift startup offers to clean homes for free—in exchange for videos of cleaners to train household robots. According to the company's calculations, the value of this data covers all service costs.
Granta Literary Prize Accidentally Published a Story Created by Artificial Intelligence
The prestigious British Commonwealth Prize discovered a story among its publications that appears to have been written by AI. Experts found typical signs of LLM-generated text.
Google AI Overviews Mistakes Search for Assistant Instructions
Google AI Overviews contains a bug: when searching for the word 'disregard,' the system shows an assistant response like 'Got it! How can I help you?' instead of search results.
Grok Isn't Needed by Government: Why Musk's Chatbot Failed in the US Market
Reuters analyzed over 400 examples of AI usage by the American government and found Grok in only three — and only for basic tasks. This is a serious blow to Musk's ambitions and the SpaceX IPO.
Spotify Studio: AI Creates Personal Podcasts from Your Listening History
Spotify launched Studio—an AI app that creates personalized podcast briefings based on your listening history and connected applications.
Spotify to Launch AI Song Remixes in Partnership with Universal Music Group
Spotify is launching a paid service for creating AI remixes and cover versions. Artists can opt out, but those who consent will receive royalties for each generated track.
Polyend Endless: Guitar Pedal That Creates Effects from Text Commands
Polyend has released the Endless guitar pedal with artificial intelligence that creates sound effects based on user text descriptions. Price — $299.
Google's New Gemini Model Creates Realistic Videos from Anything
Google has unveiled an update to Gemini that transforms photographs and text into photorealistic videos with minimal effort—once again raising questions about where the line lies between harmless fun and dangerous AI-gen
Hackers Learn to Bypass AI Chatbot Defenses Through Manipulation of Their 'Personalities'
Early AI chatbots could be tricked into violating their safety instructions with simple requests — now hackers are deploying more sophisticated techniques to exploit their vulnerabilities.
Utah approves Stratos mega data center despite environmental risks
Utah has approved the Stratos Project data center (40,000 acres) — twice the size of Manhattan. The project will consume 9 gigawatts of electricity and threaten the region’s water supply.
Google unveiled personal AI agents for email, calendar, and information retrieval
At the I/O 2026 conference, Google showed a new generation of AI agents that can work in the background and manage email, calendars, and information search. The company believes this will be useful at scale.
Google Expands SynthID: Invisible AI-Generated Content Watermarking Becomes Standard
Google launches a major expansion of SynthID and C2PA technologies to watermark AI-generated images, videos, and audio, making it easier for people to identify fake content online.
Google Built Gemini Into Search Ads for Personalized Product Descriptions
Google embedded Gemini into search ads — the system now automatically generates product descriptions and recommendations directly in Search ads.
YouTube Shorts Remix: Google Lets Users Reimagine Videos with Gemini AI
Google has launched Remix for YouTube Shorts, a feature that lets users reimagine other people's videos using Gemini Omni — change styles, add elements, and insert themselves into other clips.
Vibe coding on Android: create apps right on your phone
Google has built AI into Android for app creation. Now any user can make their own app without programming knowledge, simply by describing the idea.
DeepMind Ready to Reimagine Drug Discovery and Defeat Diseases
At Google I/O, DeepMind's CEO announced a goal to reimagine drug discovery with AI—but experts recommend staying realistic.
Google went too far with Gemini: a story of user AI fatigue
Google is aggressively embedding Gemini into every app, from Gmail to Docs. The story echoes Microsoft Copilot in Windows 11, which annoyed users with its intrusiveness.
Musk loses lawsuit against Altman: court finds claims invalid
The jury unanimously found that Musk's claims against Sam Altman and OpenAI were either barred by the statute of limitations or dismissed. The judge agreed with the verdict.
The Musk-Altman trial exposed a leadership problem in the AI industry
The trial between Musk and Altman, his fellow OpenAI co-founder, ended in Altman's favor: the jury rejected the claims in two hours, but the proceedings exposed deep trust problems in AI industry leadership.
Revamped Siri in iOS 27 will get an auto-delete option for chats
Apple is preparing a major Siri update for iOS 27 with automatic deletion of conversation history — part of the company’s strategy to make privacy a competitive advantage.
Students booed Eric Schmidt over AI optimism at commencement
At the University of Arizona commencement, students booed Eric Schmidt, former Google CEO, for a speech about AI and its potential, expressing frustration with optimism about the technology amid uncertainty in the job ma
McDonald's and Wendy's roll out AI voice assistants at the drive-thru
Major restaurant chains are starting to use AI chatbots to take orders via microphone — this is only the first step toward mass automation in the fast-food industry.
AI music has flooded streaming services, but few people listen to it
Generative AI is now actively creating music for Spotify and Apple Music, but it remains unclear whether listeners want it. The story of how a gimmick turned into a problem.
Sony defends Xperia's AI Camera Assistant against criticism over poor examples
Sony is trying to defend its AI Camera Assistant after criticism: the company clarifies that it is a guidance system for lighting and angles, not photo editing.
Demis Hassabis: the architect of Google DeepMind, the hidden figure in Musk's lawsuit
At the Musk v. Altman trial, a story surfaced of how the architect of Google DeepMind became one of the most influential figures in the rivalry between AI labs.
Claude succumbed to manipulation: researchers bypassed safeguards through flattery
Mindgard researchers obtained prohibited content from Claude through respect and flattery. This calls Anthropic's approach to AI safety into question.
Automakers Bet on AI: How LLMs Will Change the Car Development Process
Developing a new car takes five years and often becomes outdated before launch. AI promises to speed up design, simulation, and aerodynamic testing, cutting the development cycle in half.
Google, Microsoft and xAI to let the US review AI models before release
Three of the biggest AI companies have agreed to let the US government review new models before their public release. The government review program is moving from a pilot to a mandatory process.
OpenAI prepares its own smartphone for a 2027 launch
According to analyst Ming-Chi Kuo, OpenAI is developing its own smartphone powered by a custom MediaTek Dimensity 9600 chip, with mass production scheduled for early 2027.