Google DeepMind Reveals AI Agents with Human-Like Computer Skills
Google’s DeepMind has made a significant advancement in agentic AI, developing agents that can interact with computers in a manner similar to humans. This development is seen as a major step forward in creating more capable and autonomous AI systems. Researchers at the Bay Area Machine Learning Symposium also discussed the future of AI, with a focus on creating models that can learn through interaction and exploration rather than just being trained on massive datasets. The goal is to achieve systematic generalization, allowing AI to combine known elements into novel meanings and learn with significantly less data than current models require. Google DeepMind’s work on models for more intelligent robots, including the recently released Gemini Robotics 1.5 and E.R. 1.5, further illustrates this push towards more interactive and intelligent AI.
Sources:
Privacy Concerns Mount Over AI Chatbot Training Data
A recent report has highlighted that AI chatbots commonly use user conversations as training data, which is raising significant privacy concerns among users and experts. This practice is becoming more widespread as companies aim to improve the performance and accuracy of their large language models. The use of personal conversations for training purposes brings to the forefront the ongoing debate about data privacy and the ethical implications of AI development.
Sources:
Waymo to Launch Autonomous Robotaxi Service in London by 2026
Waymo has announced its intention to launch a fully autonomous ride-hailing service in London, with operations planned to begin in 2026. The company will start laying the groundwork with its fleet partner, Moove, and will work with local and national authorities to obtain the necessary permissions. This expansion builds on Waymo’s existing UK presence, which includes engineering hubs in London and Oxford. The UK government has been supportive of the introduction of driverless cars, with plans to facilitate regulated public trials for robotaxis as early as spring 2026. While the initial phases of testing may still require a human safety driver, Waymo’s goal is to offer rides with no one behind the wheel. The move is anticipated to create jobs and investment in the UK’s autonomous vehicle sector.
Sources:
Oshkosh Defense Introduces New Autonomous Military Vehicles
Oshkosh Defense introduced a new lineup of autonomous ground vehicles at the Association of the United States Army’s annual meeting. The Family of Multi-Mission Autonomous Vehicles (FMAV) is designed to support the U.S. Army’s modernization efforts, particularly in missile defense. The trio of vehicles includes light, medium, and heavy payload options, each adaptable for various missions such as mobile missile launching, counter-drone operations, and resupply. These platforms can be operated remotely, with an optional crew, or fully autonomously. The vehicles are based on existing Oshkosh platforms, which the company says will facilitate easier adaptation for Army missions.
Sources:
Serve Robotics’ Delivery Bot Named a ‘Next Big Thing in Tech’
Serve Robotics Inc. has been recognized in Fast Company’s “Next Big Things in Tech” list, winning the top spot in the Robotics and Automation category. The award highlights the company’s third-generation autonomous sidewalk delivery robot. This latest model is designed to be faster, travel farther, and operate more safely than its predecessors. Serve Robotics has already deployed 1,000 of these new robots in major U.S. cities and aims to have 2,000 in operation by the end of 2025. The company has also entered into a partnership with DoorDash to integrate its robots into their food delivery network, starting in Los Angeles.
Sources:
AiMOGA’s ‘Mornine’ Robot Showcases Autonomous Skills at Chery Conference
At the 2025 Chery Global Innovation Conference, AiMOGA Robotics’ intelligent robot, Mornine, delivered a fully autonomous presentation in seven languages. The demonstration showcased the robot’s abilities in language interaction, navigation, and multimodal execution. Mornine’s technology is based on Chery’s vehicle-engineering platform, incorporating expertise from intelligent cockpits and driver-assistance systems. The robot also performed a complete, autonomous car-purchasing reception process at the event, including greeting, vehicle explanation, and opening the car door. This event marked a significant step in the collaboration between the automotive and robotics industries.
Sources:
OpenAI and Broadcom Partner to Develop Custom AI Accelerators
OpenAI and Broadcom are collaborating on a plan to ship OpenAI-designed AI accelerators starting in 2026. The partnership will utilize Ethernet-first racks and aims to give OpenAI greater control over its capacity and costs. Broadcom will provide the network backbone, optics, and manufacturing capabilities necessary for the large-scale production. This strategic co-development is expected to integrate model lessons directly into the silicon, reducing latency across the technology stack.
Sources:
Microsoft Releases Local Azure DevOps MCP Server for AI Integration
Microsoft has announced the general availability of the Azure DevOps local MCP Server, which acts as an intermediary between AI assistants and an organization’s Azure DevOps setup. This local server allows for the injection of context from Azure DevOps, such as work items, pull requests, and test plans, into prompts for large language models (LLMs) to provide more relevant and tailored responses. A key feature of this release is its “local-first” security model, ensuring that sensitive data does not leave the user’s network or local development environment. The server supports major areas within Azure DevOps, including work items, wiki, repos, search, and test plans. Microsoft’s strategy is to support customers invested in Azure Boards and Azure Pipelines while encouraging source code migration to GitHub for enhanced agentic AI features with GitHub Copilot.
Sources:
Why GitOps is Becoming Essential for Mature DevOps Practices
GitOps is gaining traction as the next stage in the evolution of DevOps, providing a framework for consistency, compliance, and velocity in software delivery. By using Git as the single source of truth for both application and infrastructure code, GitOps extends the “infrastructure as code” paradigm to the entire system state. This approach helps platform engineering teams empower developers with standardized workflows and reusable modules while maintaining necessary governance. Organizations adopting GitOps are seeing benefits such as reduced mean time to remediation, continuous compliance through automated evidence collection, and faster delivery cycles. As AI-generated code becomes more prevalent, the importance of trustworthy automation through GitOps is expected to increase.
Open-Source Makes AI Chips More Accessible, Says Jim Keller
According to leading microprocessor engineer Jim Keller, the development of artificial intelligence chips is becoming more accessible and affordable due to open-source technology. Speaking at the GITEX Global event, Keller challenged the idea that creating AI processors requires massive investments, highlighting his company’s open-source approach. He also expressed skepticism about the long-term effectiveness of chip export restrictions, suggesting they accelerate technological independence in the restricted regions.
Sources:
Microsoft’s October 2025 Patch Tuesday Fixes 172 Flaws, Including Six Zero-Days
Microsoft has released its October 2025 Patch Tuesday security updates, addressing a total of 172 vulnerabilities. Among the fixes are patches for six zero-day vulnerabilities, some of which were being actively exploited. The updates cover a wide range of Microsoft products and include fixes for critical remote code execution and elevation of privilege flaws. This month’s security release is particularly notable as it coincides with the end of the support lifecycle for Windows 10.
Sources: