1. OpenAI Launches ChatGPT Images 2.0 with Advanced Thinking Mode
OpenAI has officially rolled out ChatGPT Images 2.0, its new and improved image generation model, to all users. This latest version introduces a unique 'Thinking mode', which enables more sophisticated and advanced image creation capabilities. Users can now expect enhanced performance and greater control over their generated visuals.
Read full story2. Claude Code Now Offers Quick Session Recaps
Claude Code has introduced a new feature that provides users with a concise, one-line summary of their session activity. This recap appears automatically after three minutes of inactivity, aiming to help users quickly catch up on their work or context without manually searching through past interactions. This update streamlines the user experience for developers and others utilising the Claude Code platform.
Read full story3. Cowork Introduces New Live Artifacts Tab
Cowork has recently launched a new 'Live Artifacts' tab, enhancing its collaborative platform. This new feature allows users to directly view and manage dashboards and trackers that are linked to their files within Cowork, streamlining workflows and improving project oversight.
Read full story4. Google Upgrades Deep Research with Gemini 3.1 Pro Enhancements
Google has announced significant upgrades to its Deep Research platform with the introduction of Deep Research Max, powered by Gemini 3.1 Pro. This enhanced version now offers advanced functionalities including MCP (Multi-modal Content Processing) connections, allowing for richer data integration. Users will also benefit from inline charts for immediate data visualisation and direct access to AI Studio, providing a comprehensive environment for advanced AI-driven research.
Read full story5. Google AI Plans Now Include AI Studio Access for Developers
Developers subscribing to Google AI Pro and Ultra plans will now automatically receive access to AI Studio. This integration aims to streamline the development process and provide enhanced tools for AI model creation and deployment within Google's ecosystem. The bundle offers a comprehensive suite of AI capabilities for advanced projects.
Read full story6. Codex Achieves 4 Million User Milestone
Codex, the AI coding assistant, has successfully surpassed the significant milestone of 4 million active users. To express gratitude to its growing community, the platform has reset its rate limits, allowing users greater access and flexibility. This achievement highlights Codex's increasing popularity and impact within the developer community.
Read full story7. Codex Resets Rate Limits for 4 Million Users
Codex has announced the reset of its rate limits, effective today. This action is a gesture of gratitude to the community for helping the platform achieve a significant milestone of 4 million users. The company implemented this change to acknowledge the support and engagement of its growing user base.
Read full story8. Jensen Huang Explains NVIDIA's Enduring Advantage
NVIDIA CEO Jensen Huang elaborates on the company's competitive "moat," clarifying that it extends much further than just their renowned semiconductor chips. He specifies that NVIDIA's strength lies in its comprehensive five-layer technology stack, which integrates hardware, software, and services, offering a complete ecosystem that is difficult for competitors to replicate. This integrated approach, rather than individual components, secures NVIDIA's market position.
Read full story9. OpenAI Builds 'Always-On' Agent Platform for ChatGPT
OpenAI is reportedly developing a new platform for ChatGPT that will enable "always-on" capabilities for its AI agents. This innovative platform aims to provide various skills and scheduled triggers, allowing ChatGPT to perform persistent tasks and actions without constant human prompting. This development signifies a major step towards more autonomous and proactive AI assistants.
Read full story10. Google Develops AI Tour Guide for Street View
Google has launched "Talking Tours," an innovative AI tour guide integrated with Street View. This technology enriches virtual exploration by describing monuments and points of interest. Users can ask questions directly to the AI, receiving immediate and informative responses, thereby enhancing their understanding and experience of various locations worldwide.
Read full story11. WildDet3D: A New Model for Monocular 3D Object Detection
WildDet3D introduces an innovative promptable 3D object detection model designed specifically for in-the-wild monocular 3D detection. This advanced model capabilities allow it to accurately identify and locate 3D objects using only a single 2D image, even in complex and unconstrained environments. Researchers developed WildDet3D to overcome limitations of traditional methods, offering higher precision and adaptability in real-world scenarios.
Read full story12. CubeSandbox Powers AI Agents with Rapid VM Cold Starts
CubeSandbox is an innovative sandbox environment designed to provide hardware-isolated virtual machines (VMs) specifically for AI agents. This platform boasts the impressive capability of initiating these VMs from a cold start in less than 60 milliseconds. This rapid deployment feature significantly enhances the performance and responsiveness of AI applications by reducing latency and ensuring a secure, isolated operational environment for artificial intelligence development and deployment.
Read full story13. Transform Code Repositories into Interactive Courses
The "codebase-to-course" tool helps developers effortlessly convert any existing code repository into a comprehensive, self-contained educational course. This innovative solution allows users to create interactive learning experiences complete with structured modules and engaging quizzes, making it ideal for onboarding new team members or teaching programming concepts.
Read full story