Reporting AI nonsense. A future news media, driven by virtual assistants 🤖
Devstral from Mistral AI tops open-source benchmarks for agentic tasks
Mistral AI and All Hands AI launched Devstral, a 24B-parameter open-source language model for software engineering, outperforming key benchmarks. It runs locally, supports 128K tokens, and is available under Apache 2.0 via multiple platforms.
🗞 #mistral
Comet enters early testing as Perplexity debuts its agentic browser
Perplexity's Comet Browser integrates its AI agent directly into core browsing functions, enabling autonomous tab management, task execution, and contextual content understanding—positioning it as a full-stack assistant for advanced workflows.
🗞 #perplexity
Google launches coding agent Jules in beta with free daily tasks
Google's coding agent Jules is now in global beta, letting developers assign tasks directly from GitHub issues. Powered by Gemini 2.5 Pro, it automates pull request workflows and includes a free tier of five tasks per day.
🗞 #ai
First look into upcoming AI-generated Video Overviews from Google
Google's Illuminate project hints at a major shift toward AI-generated multimedia, including audio summaries of texts and short, fully generated video overviews from prompts, powered by a unified model likely tied to Gemini or Veo technologies.
🗞 #aitestkitchen
Google prepares to launch Flow, a new video editing tool, at I/O 2025
Google is set to introduce Flow, a video generation tool likely debuting at Google I/O, built on updated Veo and Imagen models. Flow appears to revive the Storyboard concept, guiding users from prompt to structured video output.
🗞 #aitestkitchen
Windsurf launches SWE-1 AI model for real-time on-device use
Windsurf's Wave 9 SWE-1 is a transformer-based AI model optimized for low-latency, on-device use without cloud dependency, offering real-time multimodal capabilities tailored for privacy-focused, mobile-first applications.
🗞 #windsurf
Microsoft tests in-app shopping with Copilot checkout system
Microsoft is developing a native checkout system within Copilot, aiming to integrate platforms like Shopify and shift toward in-app commerce. This aligns with its broader strategy to position AI assistants as transaction tools, not just search aids.
🗞 #microsoftcopilot
OpenAI prepares SWE Agent that answers code questions and drafts PR
OpenAI is previewing a "Software Engineering Agent" within ChatGPT, designed to answer code queries, run code, and draft pull requests. The tool is expected to integrate deeply into the IDE workflow, beginning with desktop deployment.
🗞 #chatgpt
Video Overviews settings spotted in NotebookLM ahead of I/O
Google is expanding NotebookLM with a video overviews feature that generates podcast-style summaries using visuals and voice. Users can customize format, length, and content focus. A launch is likely at the upcoming Google I/O event.
🗞 #notebooklm
ChatGPT may soon support third-party integrations via MCPs
OpenAI is testing a "custom connection" feature in ChatGPT based on the Model Context Protocol, allowing bespoke integrations with third-party tools. This move aligns with its focus on workplace productivity and agent-based workflows.
🗞 #chatgpt
New Claude Neptune model undergoes red team review at Anthropic
Anthropic is testing a new AI model, Claude Neptune, with internal safety reviews and red team evaluations targeting its constitutional AI system. Positioned for release soon, it may offer architecture changes aimed at secure, high-performance applications.
🗞 #claude
New Comet browser by Perplexity may post to X and browse internet for you
Perplexity’s upcoming Comet browser is positioned to function as an autonomous web agent, capable of executing authenticated online tasks. Its native automation could set a new standard in AI-native browsing ahead of competitors like Google and Dia.
🗞 #perplexity
Gemini Ultra poised to compete with GPT-5 in multi-modal AI race
Google is updating Gemini with new multi-modal tools like Veo 2 and Memory, expanding its Agents toolbox. A Gemini Ultra tier will bundle advanced features, supporting a tiered model ahead of competitive AI launches at the upcoming Google I/O.
Google is also working on a possibility to upload files in Deep Research. Soon, you will be able to upload your documents and images to be used as a reference for Deep Research.
🗞 #gemini
Google is developing Video Overviews feature for NotebookLM
Google is preparing to add video overviews to NotebookLM, using its Veo 2 model to summarize notebook content visually. A new “Editor’s Picks” section and a mobile app launch signal a shift toward public content sharing and cross-platform access.
🗞 #notebooklm
OpenAI released reinforcement fine-tuning with o4-mini, announced during the 12 Days of OpenAI in December!
Читать полностью…Claude Sonnet 4 and Opus 4 spotted in early testing round
Anthropic is nearing launch for its Claude 4 model family, currently in limited internal testing. Claude Sonnet 4 and Claude Opus 4 are under evaluation with strict access limits and higher safety classification due to advanced capabilities.
🗞 #claude
Google expected to add credit system to Flow AI video editor
Google's Flow is a modular video creation tool in early testing, offering text, frame, and asset-based inputs. Powered by Veo models and Gemini, it uses a credit system, suggesting a move toward a structured, creator-focused platform.
🗞 #aitestkitchen
Google launches NotebookLM mobile app with audio-first features on mobile
Google is launching the NotebookLM mobile app, bringing its web-based features to iOS and Android with a focus on audio content. The app supports podcast-style summaries, interactive playback, and flexible knowledge consumption on the go.
🗞 #notebooklm
Google readies upgrade to Stream Realtime feature in AI Studio
Google quietly updated AI Studio, hinting at upcoming support for real-time multimodal processing via a potential Flash 2.5 model. Backend changes and developer polls suggest an evolving agent system for coding and deployment via Cloud Run.
🗞 #aistudio
OpenAI rolled out Codex for automated coding tasks in ChatGPT Pro
OpenAI's Codex is a cloud-based software agent for automating coding tasks like bug fixes and feature implementation. Available May 16, 2025, for select ChatGPT tiers, it operates securely within user codebases with verifiable outputs.
🗞 #chatgpt
Microsoft begins testing user memory feature in Copilot for Pro users
Microsoft is testing a personalization feature in Copilot that allows the tool to retain user context, similar to ChatGPT’s memory. Currently limited to some Pro users, it reflects Microsoft’s gradual push toward deeper AI integration.
🗞 #microsoftcopilot
Sketchpad and native Mermaid support coming soon to Grok
Grok introduces Sketchpad for direct hand-drawn input and integrates native Mermaid diagram support, allowing users to generate and view structured visuals within the interface, targeting technical workflows across planning and development tasks.
🗞 #grok
Grok set to gain Tasks feature for periodical execution
xAI is developing a “Tasks” feature for Grok, allowing users to automate recurring prompts with detailed scheduling. If output delivery expands beyond the app, Grok could shift toward a practical assistant model with broader utility.
🗞 #grok
Gemini set for major upgrade with 7 features debuting at Google I/O
Google is restructuring its Gemini AI subscription tiers, introducing Gemini Pro and Ultra with advanced features like image generation and Drive perks. New tools and UI updates signal a shift toward enterprise and integrated AI solutions.
🗞 #gemini
Perplexity tests new Projects feature for building apps and documents with AI
Perplexity Projects, currently in beta on iOS, enables users to build and share mini web apps or structured documents using AI-driven research and code generation, positioning the platform as a tool for development and prototyping.
🗞 #perplexity
Manus AI now provides free daily credits to all users which can cover one task. That’s a good change to help ppl to get onboard.
Читать полностью…ChatGPT could gain PDF downloads and new third-party connectors
OpenAI is advancing ChatGPT with expanded third-party connectors, PDF export options, and a possible usage-based credit system, signaling a shift toward modular, scalable enterprise applications and broader productivity tool integration.
🗞 #chatgpt
Google to unveil Imagen 4, Imagen 4 Ultra and Veo 3 models at I/O 2025
Google is set to unveil new versions of its Veo and Imagen models at the May 20 I/O conference, signaling advances in generative media. These updates suggest expanded capabilities across video, image synthesis, and multimodal applications.
🗞 #gemini
Perplexity is developing natural language tab control for Comet browser
Perplexity is advancing its AI browser Comet with natural language controls for managing tabs and history. New features support multi-tab commands, tab grouping, and contextual search, pointing toward a conversational browsing model.
🗞 #perplexity
Microsoft released Pages for Copilot 🔥
"Meet Pages, your new editing BFF! No need to start from scratch - just hit "Edit this Response," highlight text, and tap "Ask" to tweak, expand, or polish your writing."