AI Adoption in US Workplaces Surges, Gallup Poll Finds

A recent Gallup Workforce survey reveals a significant increase in the adoption of artificial intelligence among American workers. The poll, which surveyed over 22,000 U.S. workers, found that 12% of employed adults now use AI daily in their jobs. Approximately one-quarter of respondents reported using AI frequently (at least a few times a week), while nearly half of the workers surveyed stated they use AI at least a few times a year. This marks a notable rise from 2023, when just 21% of workers reported using AI at least occasionally, highlighting the rapid integration of generative AI tools like ChatGPT into professional environments.

Sources:

NTSB Investigates Waymo Robotaxis for Illegally Passing School Buses

The U.S. National Transportation Safety Board (NTSB) has launched an investigation into Waymo, a subsidiary of Alphabet, following reports of its autonomous robotaxis illegally passing stopped school buses in Austin, Texas. The incidents occurred on at least 19 occasions since the beginning of the school year, raising significant autonomous vehicle safety concerns. In December, Waymo recalled over 3,000 vehicles to update the software responsible for the issue. Despite the updates, the Austin Independent School District reported five more incidents in November and requested Waymo suspend operations near schools during peak hours, a request Waymo reportedly declined. Waymo stated that no collisions occurred and expressed confidence that its autonomous driving system’s performance around school buses is superior to that of human drivers.

Sources:

Airbus to Test Humanoid Robots in Aircraft Manufacturing

Aerospace giant Airbus has partnered with Chinese robotics company UBTech to explore the use of humanoid robots in its aircraft manufacturing processes. The agreement involves Airbus testing UBTech’s Walker S2 industrial humanoid robot for early-stage factory automation concepts. This initiative marks a new phase in Airbus’s automation strategy, which has historically focused on fixed and mobile robotic systems for tasks like drilling and inspection. Humanoid robots are being considered for their potential to work in environments designed for humans, complementing existing automation. UBTech anticipates its production capacity for industrial humanoid robots will exceed 10,000 units in 2026.

Sources:

iRobot Emerges from Bankruptcy as Picea U.S. Subsidiary

Following a successful restructuring, iRobot has emerged from Chapter 11 bankruptcy. The company now operates as a U.S. subsidiary of the Chinese manufacturer Picea. As part of its reorganization, iRobot has also established a new data security unit in the United States.

Sources:

U.S. Marine Corps Designates GenAI.mil as Official Enterprise AI Platform

The U.S. Marine Corps has officially named GenAI.mil as its preferred enterprise platform for large language models (LLMs), authorizing its use for all Marines, civilians, and contractors. This decision is part of a broader Department of Defense (DoD) initiative to accelerate the adoption of generative AI across the military. According to Captain Christopher Clark, the service’s AI lead, GenAI.mil represents the first true enterprise solution for Marines to utilize generative AI. The platform serves as the Pentagon’s central hub for commercial LLMs, with Google’s Gemini being the first model integrated.

Sources:

The Hidden Costs of DevOps and SaaS Downtime in Cloud-First Businesses

In 2024, popular DevOps SaaS platforms like GitHub, Jira, and Azure DevOps faced 502 incidents, resulting in over 4,755 hours of degraded performance and outages. This highlights the significant financial risks for cloud-first businesses, where downtime costs can be substantial. For Fortune 1000 companies, the hourly cost of an outage can range from $1 million to over $5 million. To mitigate these risks, businesses are advised to adopt a multi-layered data protection strategy, including frequent backups, immutable storage, and continuous recovery plan testing. A growing number of UK businesses are re-evaluating their cloud strategies due to rising costs, opting for hybrid models and repatriating workloads to private clouds or on-premise deployments to prioritize data locality and sovereignty.

Sources:

Kubernetes 1.32 and 1.33 Updates: Enhanced Security and AI/ML Integration

The Kubernetes development community has been active, with the 1.36 release cycle set to begin soon. A recent key change includes the default enabling of the WatchCacheInitializationPostStartHook to improve watch cache reliability. The Kubernetes 1.32 release introduced significant security enhancements, including improved Pod Security Standards and native secret encryption at rest with automatic key rotation. Looking ahead, the Kubernetes 1.33 release, expected in mid-2026, will focus heavily on AI and Machine Learning integration, featuring advanced GPU resource management and native support for distributed training frameworks.

Sources:

Microsoft Enhances Quantum Development Kit with AI-Assisted Programming

Microsoft has released a significant update to its Quantum Development Kit (QDK), integrating AI-powered features to streamline the development of quantum applications. The upgraded open-source toolkit is designed to help researchers transition from noisy physical qubits to more stable logical qubits. Key new features include GitHub Copilot integration for AI-assisted coding, pattern suggestions, and task automation. The QDK now runs on standard laptops and integrates with Visual Studio Code, lowering the barrier to entry for quantum development. To accelerate research, the kit includes new libraries for chemistry and error correction and supports multiple frameworks, including Q#, OpenQASM, and Qiskit.

Ramp’s Internal AI Coding Agent Drives 30% of Pull Requests

Fintech company Ramp has developed an internal AI coding agent named Inspect, which is already responsible for approximately 30% of merged pull requests in its primary repositories. This advanced tool provides the AI agent with the same level of access to the development environment as a human engineer. Operating within sandboxed virtual machines, Inspect integrates with databases, CI/CD pipelines, monitoring tools, and platforms like Slack and GitHub. This comprehensive access allows the agent to not only write code but also to verify its functionality through testing and monitoring, addressing a common limitation in many AI coding assistants. The system uses Cloudflare Durable Objects for state management, enabling the agent to maintain context across interactions.