News/New Launches
Microsoft launches CoreAI division to accelerate AI infrastructure development and integration
Microsoft launched a significant new AI division called CoreAI - Platform and Tools to accelerate AI infrastructure development and integration across its software ecosystem. Strategic overview: Microsoft aims to fundamentally transform application development by integrating autonomous AI agents into all software categories by 2025, representing a dramatic acceleration of technological advancement. The newly formed CoreAI division combines expertise from Microsoft's Developer Division, AI Platform, and Office of the CTO teams Jay Parikh will lead the initiative as Executive Vice President CEO Satya Nadella describes this effort as compressing "30 years of change into 3 years" for application development Technical infrastructure...
read Jan 16, 2025Google’s new LLM architecture cuts costs with memory separation
Large language models (LLMs) are getting a significant upgrade through Google's new Titans architecture, which reimagines how AI systems store and process information by separating different types of memory components. Key innovation: Google researchers have developed a neural network architecture called Titans that extends model memory capabilities while keeping computational costs manageable. The architecture introduces a novel three-part system that handles information processing and storage differently from traditional LLMs By segregating memory functions, Titans can process sequences up to 2 million tokens in length Early testing shows Titans outperforming GPT-4 on long-sequence tasks despite using fewer parameters Technical framework: The Titans architecture...
read Jan 16, 2025Luma Labs ray 2 AI-powered video generator rivals Sora
Luma Labs has launched Ray 2, an AI-powered video generation tool that can create realistic 10-second clips from text descriptions on its Dream Machine platform. Key Features and Capabilities: Ray 2 represents a significant advancement in AI video generation, powered by 10 times more computational resources than its previous version. The system can generate realistic video clips featuring complex scenarios with natural motion and physics Users can access the platform through both free (720p resolution) and premium tiers (1080p resolution at $66.49 monthly) The tool focuses on user-friendly operation and quick generation times Platform Accessibility and Future Development: Dream Machine serves...
read Jan 16, 2025Google Chrome is testing a new AI feature that combats annoying pop-up ads
A new Google Chrome feature called PermissionsAI is being tested to reduce intrusive website permission pop-ups by using artificial intelligence to predict user preferences. The core functionality: Google's PermissionsAI leverages the company's Permission Predictions Service and Gemini Nano v2 to anticipate whether users would approve or deny website requests for notifications and location sharing. Rather than displaying disruptive pop-ups, the system moves likely-to-be-denied requests to a more subtle interface The feature analyzes past user behavior patterns to make informed predictions about future responses It works alongside Chrome's Safe Browsing security tool to maintain protection against malicious sites Technical implementation: The...
read Jan 16, 2025How startup Ndea aims to build self-improving AI systems
François Chollet, creator of Keras, and Mike Knoop, Zapier co-founder, have launched Ndea, a new AI research lab focused on developing artificial general intelligence (AGI) through a combination of deep learning and program synthesis. The core innovation: Ndea aims to merge intuitive pattern recognition with formal reasoning through "guided program synthesis," moving beyond traditional deep learning's limitations. The approach combines deep learning's pattern recognition capabilities with program synthesis, which searches for discrete programs that explain data This hybrid method could enable AI systems to generalize and learn from fewer data points than current approaches The company derives its name from...
read Jan 16, 2025Luma Labs launches new AI video model with improved motion and physics
Luma AI has unveiled Ray2, its latest generative video model, offering enhanced motion and physics capabilities through its Dream Machine platform. What's new: Ray2 represents a significant advancement in AI video generation, trained with 10 times more computing power than its predecessor. The model specializes in creating 5-10 second video clips from text descriptions Generation times can vary from seconds to minutes depending on user demand The platform maintains tiered pricing, from free 720p generations to enterprise plans at $1,672.92 annually Technical capabilities: The new model demonstrates improved capabilities in generating natural movement and realistic physics simulations. Early tests show...
read Jan 16, 2025Gmail’s new button automates AI responses from Gemini
Google has introduced a new "Insert" button for its Gemini AI email feature in Gmail on Android, streamlining the process of using AI-generated email responses. New feature overview: Google's latest update to Gmail on Android introduces a simple yet impactful enhancement to its Gemini AI integration system. The new "Insert" button allows users to directly transfer AI-generated responses into Gmail's reply window The button appears at the bottom right of the screen, alongside thumbs up/down options and below formatting tools Users can edit the inserted text before sending Current functionality: Gemini's AI capabilities in Gmail extend well beyond basic email...
read Jan 16, 2025The Halliday AI smart glasses might be the most innovative wearable yet
The breakthrough: Gyges Labs has introduced AI-enabled smart glasses featuring a unique top-mounted display module that projects images directly to the wearer's eye, departing from traditional lens-integrated displays. The glasses feature a circular monochrome display in black and green colors, reminiscent of the interface from the Fallout video game series The display projects what appears similar to a smartwatch interface, though company representatives compare it to viewing an iPad from a meter away A color display version is planned for the next generation of the product [embed]https://www.youtube.com/watch?v=4vJf7gmNGGI[/embed] Technical specifications and features: The Halliday AI Glasses combine hardware innovation with AI-powered...
read Jan 16, 2025Samsung’s latest Galaxy S25 AI teaser hints at advanced virtual assistant
Samsung has just released a new teaser showcasing advanced AI capabilities coming to their Galaxy S25 smartphone series, with features centered around a more conversational and integrated digital assistant experience. Key Features Overview: The new Galaxy AI system promises seamless cross-app functionality and natural language interactions, representing a significant upgrade to Samsung's existing Bixby assistant. The teaser demonstrates a user effortlessly finding a pet-friendly Italian restaurant with outdoor seating, sharing the details with contacts, and adding the event to their calendar in one continuous interaction The system appears capable of understanding context and executing complex multi-step tasks across different applications...
read Jan 16, 2025MiniMax releases new open-source LLM with 4M token context window
MiniMax, a Singaporean AI company, has released and open-sourced a new family of AI models featuring an unprecedented 4-million token context window, doubling the previous industry record. Key innovation: MiniMax's new language model series introduces groundbreaking context handling capabilities that allow it to process the equivalent of a small library's worth of text in a single exchange. The MiniMax-01 series includes both a foundation large language model (MiniMax-Text-01) and a visual multi-modal model (MiniMax-VL-01) The models are now available through Hugging Face, Github, Hailuo AI Chat, and MiniMax's API Pricing is highly competitive at $0.2 per million input tokens and...
read Jan 15, 2025Microsoft autogen update boosts AI agent flexibility and programming language support
Microsoft has released AutoGen v0.4, an updated orchestration framework that enhances AI agent flexibility and organizational control through improved modularity, cross-language support, and enhanced observability features. Key updates and features: The new version addresses previous architectural constraints and inefficiencies while introducing critical improvements to the framework's functionality. Asynchronous messaging capabilities now support event-driven and request-interaction patterns A more modular design enables plug-in components and long-running agents Cross-language interoperability currently supports Python and .NET, with more languages planned Built-in metric tracking, message tracing, and debugging tools enhance agent monitoring Framework architecture: Microsoft has implemented a three-tiered structure to better organize system...
read Jan 15, 2025LinkedIn’s new AI tool aims to match you with the perfect job
LinkedIn introduces AI-powered Job Match feature to streamline job search and recruitment by providing personalized insights about how well candidates align with open positions. The core functionality: Job Match analyzes user profiles against job listings to evaluate skills and experience alignment, helping job seekers target opportunities where they're most likely to receive responses. The feature provides instant feedback on profile-to-job compatibility using advanced in-house language models Job Match insights are available for all LinkedIn job listings Different levels of match analysis are provided based on user account type Premium features and AI integration: LinkedIn Premium subscribers receive enhanced Job Match...
read Jan 15, 2025An inside look at the autonomous supercars debuted by Maserati at CES 2025
Maserati showcased two autonomous MC20 supercars at CES 2025, including a self-driving Coupé model that recently achieved a record-breaking speed of 177 mph without human intervention. The technology breakthrough: The autonomous MC20 models represent a significant advancement in high-speed self-driving capabilities, developed through the Artificial Intelligence Driving Autonomous (AIDA) project at Milan Polytechnic. The MC20 Coupé served as the safety car at the 2025 CES Autonomous Challenge in Las Vegas A convertible MC20 Cielo version featuring the same autonomous technology was unveiled at the main CES show The project previously made history when an autonomous MC20 became the first self-driving...
read Jan 15, 2025Meta’s new AI model can translate speech from 100+ languages
Meta has unveiled SeamlessM4T, a new AI model capable of translating speech across 101 languages, marking significant progress toward real-time language interpretation technology. Key innovation: Meta's SeamlessM4T model enables more direct speech-to-speech translation, improving upon traditional multi-step approaches that convert speech to text, translate the text, and then convert it back to speech. The model demonstrates 23% higher accuracy in text translation compared to leading existing systems While Google's AudioPaLM can handle 113 languages, it only translates into English, whereas SeamlessM4T can translate into 36 different languages The technology leverages parallel data mining to match audio with subtitles from web...
read Jan 15, 2025Google is making AI features free in Gmail and Docs — but increasing its Workspace prices
Google has expanded access to AI features across its Workspace suite while simultaneously increasing base subscription costs. The core change: Google is now providing its AI features in Gmail, Docs, Sheets, Meet, and other Workspace applications at no additional cost, eliminating the previous $20 per user monthly fee for the Gemini Business plan. The base subscription price for Workspace is increasing from $12 to approximately $14 per user per month Users gain access to the Gemini chatbot, which offers advanced search and information retrieval capabilities across applications The change reflects Google's strategy to compete with Microsoft, OpenAI, and others in...
read Jan 15, 2025Microsoft relaunches Copilot for business, now with pay-as-you-go AI agents
Microsoft has launched Microsoft 365 Copilot Chat, a free business AI chat service with pay-as-you-go AI agents, aiming to drive adoption of its premium Copilot offerings. Key Features and Functionality: Microsoft 365 Copilot Chat represents a rebranding of Bing Chat Enterprise and now includes access to AI agents within the chat interface. The platform utilizes GPT-4 for queries and allows users to upload files for analysis, including Word documents and Excel spreadsheets Users can create and employ AI agents through Copilot Studio, accessing both web data and work-related information via Microsoft Graph The service maintains the core chat experience while...
read Jan 15, 2025Opera adds AI-powered widget to iOS browser
Opera has released a major iOS update featuring new AI-focused widgets and expanded access to its AI assistant Aria for iPhone and iPad users. What's new: Opera's latest iOS browser update introduces customizable Home Screen widgets and makes its AI assistant Aria available to all users, even without signing in. New widgets provide one-tap access to key features including Private Mode browsing and Aria AI interactions Long-press functionality on widgets reveals additional shortcuts The update removes the login requirement for using Aria AI features AI capabilities and limitations: Aria offers both text and voice-based interactions while incorporating advanced image generation...
read Jan 15, 2025DJI’s newest drone collapses into a truly portable size and comes with AI features
DJI has unveiled its latest drone, the Flip, featuring a compact foldable design and advanced camera capabilities targeting the vlogger market. Product Overview: The DJI Flip introduces a unique combination of portability and professional-grade features in a sub-250g drone platform. The drone incorporates a 48MP CMOS sensor (1/1.3-inch) capable of recording 4K HDR video at 60FPS Slow-motion capabilities allow for 4K recording at 100FPS Flight time extends beyond 30 minutes on a single battery charge Six AI-powered shooting modes enable autonomous self-filming capabilities [caption id="" align="alignnone" width="666"] Image credit: Android Authority[/caption] Design Innovation: The Flip's standout feature is its revolutionary...
read Jan 15, 2025Google Cloud’s new AI platform equips retailers with revolutionary new tools
Google Cloud has unveiled a comprehensive suite of AI-driven retail solutions ahead of NRF 2025, designed to address industry challenges including rising costs, supply chain issues, and customer personalization demands. Core technology introduction: Google Agentspace will be the centerpiece platform, combining AI agents with Gemini's reasoning capabilities and enterprise-grade search functionality. The platform enables retail employees to use natural language queries for accessing product and operational data Vertex AI Agent Builder allows retailers to create custom AI agents for specific needs NotebookLM for enterprise focuses on improving employee productivity through enhanced data access Search capabilities enhancement: Vertex AI Search for...
read Jan 15, 2025Google’s Gemini AI can now process video and image simultaneously — here’s why it’s important
Google's Gemini AI has achieved simultaneous processing of multiple visual streams through an experimental application called AnyChat, marking a significant advancement in AI visual processing capabilities. The breakthrough explained: Through AnyChat, Gemini can now process live video feeds and analyze static images at the same time, a capability previously thought impossible in AI systems. AnyChat creator Ahsen Khaliq revealed that this functionality exists within Gemini's API but hasn't been implemented in Google's official applications The system can maintain conversational coherence while handling multiple visual inputs simultaneously This capability stands in contrast to other AI platforms like ChatGPT, which must disable...
read Jan 14, 2025Wyze’s new AI-powered cameras turn security footage into text notifications
Wyze has launched an AI-powered feature that generates detailed text descriptions of security camera footage for its users. What's new: Wyze's Descriptive Alerts feature provides detailed text summaries of motion events captured by its security cameras, going beyond basic motion detection notifications. The system can describe specific details like colors and actions, such as noting the color of a delivery person's hat or describing vehicles in view Alerts provide contextual information about events occurring in the camera's field of view Users receive these descriptions as notifications, reducing the need to manually review security footage Service details and availability: The new...
read Jan 14, 2025Samsung’s Sketch-to-Image feature will soon let you incorporate voice commands too
One of Samsung's next big reveals will reportedly include a new AI-powered Sketch to Image tool that allows users to combine voice and text inputs with drawings to generate images. Latest Features: Samsung is expanding the capabilities of its Galaxy AI-powered Sketch to Image tool within Drawing Assist, introducing multimodal input functionality that enhances creative possibilities. Users can now complement their sketches with typed text or voice commands to generate more detailed and specific images The upgraded tool allows for intuitive combinations, such as sketching a cat and adding "spacesuit" via text or voice to create a composite image Previously...
read Jan 14, 2025ChatGPT now handles reminders and to-do lists — plus more on its forthcoming agents
OpenAI is introducing Tasks, a beta feature in ChatGPT that enables users to schedule future actions and reminders, marking the chatbot's expansion into digital assistant territory. Core functionality and access: ChatGPT's Tasks feature allows subscribers to schedule one-time or recurring actions, representing a significant evolution from its real-time-only interactions. The feature is currently available exclusively to Plus, Team, and Pro subscribers Users can access Tasks by selecting "4o with scheduled tasks" in the model picker A limit of 10 active tasks can run simultaneously Task management is available through chat threads or a dedicated Tasks section on the web interface...
read Jan 14, 2025Pebble Beach introduces AI cameras and shot tracing tech to its nine-hole golf course
Pebble Beach Resorts has introduced ReelGOLF®, an AI-powered video capture system, at The Hay's ninth hole, allowing golfers to create and share high-quality videos of their shots. What's new: The ReelGOLF Experience combines AI-enabled cameras, shot tracing technology, and social sharing capabilities to enhance the golfing experience at the Tiger Woods-designed short course. AI-powered Bosch cameras positioned near tee boxes and greens automatically record players' shots Players activate the system by scanning a QR code, receiving enhanced videos with shot tracing and branding within minutes Starlink satellite internet ensures reliable video delivery even in remote course locations Technology implementation: The system...
read