Weekly Tech News: AI, Robotics, Autonomous Vehicles & Cloud Innovations

Google’s Gemini 2.5 AI Wins Gold Medal in Global Programming Contest

A specialized version of Google DeepMind’s Gemini 2.5 AI model, known as Deep Think, has achieved a gold medal-level ranking at the prestigious International Collegiate Programming Contest (ICPC). Held in Baku, Azerbaijan, the competition featured 139 of the world’s top university programming teams. Gemini 2.5 Deep Think successfully solved 10 of the 12 complex algorithmic problems, placing it second overall compared to the human teams.

Significantly, the AI solved one particularly difficult problem that no human contestant could. This achievement is being hailed as a historic AI milestone, comparable to Deep Blue’s chess victory and AlphaGo’s success in Go. Google DeepMind considers this a major step forward in abstract problem-solving, with potential applications in fields like drug design and semiconductor development.

Figure AI Raises $1B+ to Scale Humanoid Robot Production with Nvidia

Figure AI, a company specializing in humanoid robots, has raised over $1 billion in its Series C funding round, elevating its valuation to $39 billion. The round was led by Parkway Venture Capital and saw major investments from Nvidia, Brookfield Asset Management, and Intel Capital.

The new capital will accelerate the deployment of Figure’s general-purpose humanoid robots into real-world applications. A large portion of the funds is earmarked for scaling up manufacturing and expanding the company’s GPU-based infrastructure for training its AI models. Figure’s proprietary AI platform, Helix, is crucial for the robots’ ability to perceive, reason, and interact with their environment. The partnership with Nvidia will be instrumental in building the next-generation GPU infrastructure required for advanced simulation and AI model training.

Tesla Expands Autonomous Vehicle Testing to Arizona Roads

Tesla has received regulatory approval from the Arizona Department of Transportation to commence testing its autonomous vehicles on public roads. This makes Arizona the fourth U.S. state to grant Tesla permission for its autonomous vehicle program, joining Texas, California, and Nevada.

The expansion is a key part of CEO Elon Musk’s strategy to offer Robotaxi services to approximately half of the U.S. population. While specific details of the Arizona testing are still emerging, Tesla’s current Robotaxi trials in other states have operated with a “Safety Monitor” present inside the vehicles.

Abu Dhabi to Showcase Smart Mobility with ‘Autonomous Systems Week’

Abu Dhabi is set to host the inaugural Abu Dhabi Autonomous Systems Week from November 10-15, 2025. The event will highlight the emirate’s strategy for shaping the future of autonomous and smart technologies. It will feature the Abu Dhabi Autonomous Systems Summit, the DRIFTx Exhibition for autonomous solutions, and the second edition of the Abu Dhabi Autonomous Racing League (A2RL).

The emirate’s ongoing pilot projects include a Robotaxi service that has already completed over 40,000 safe journeys, covering 600,000 kilometers. Officials emphasize that autonomous systems are vital for a more sustainable transport ecosystem, offering the potential to significantly reduce carbon emissions, fuel consumption, and travel times.

YouTube Launches Generative AI ‘Dream Screen’ for Shorts Creators

YouTube has announced a new suite of generative AI tools for its Shorts creators, aiming to compete more effectively with platforms like TikTok and Instagram Reels. A standout feature is “Dream Screen,” which allows users to generate custom video backgrounds from simple text prompts.

The new tools will also include AI-powered video editing, automatic effects generation, and music syncing suggestions to streamline content creation. For instance, creators can generate unique backgrounds or let AI match video clips with trending audio. These tools are part of Google’s broader strategy to integrate generative AI across its products, lowering the entry barrier for new creators and expanding creative options for established ones. The features are expected to roll out in the coming months.

Autodesk Unveils AI Assistant and Neural CAD Models for 3D Design

Autodesk has revealed new generative AI tools for its 3D modeling and animation software, Maya, and announced the development of neural CAD foundation models. Unveiled at Autodesk University 2025, the new tools for Maya include the ‘Autodesk Assistant,’ an AI assistant that automates tasks using natural-language commands, and ‘FaceAnimator,’ a tool for generating facial and lip-sync animation from audio files. FaceAnimator can be trained on a project’s existing animation to maintain a consistent artistic style.

Separately, Autodesk introduced its upcoming neural CAD foundation models for Fusion and Forma, a new category of generative AI for design and manufacturing. These models are designed to understand physical systems and digital geometry, enabling them to automate repetitive tasks and suggest novel design alternatives. The first applications will be ‘Neural CAD for Buildings’ for architects and ’neural CAD for geometry’ for creating designs from text prompts.

Microsoft Azure Logic Apps Now Support AI Agent Integration via MCP

Microsoft has announced a public preview that allows Azure Logic Apps (Standard) to function as Model Context Protocol (MCP) servers. This update enables developers to expose existing workflows as tools that can be used by large language models and AI agents.

The new capability is designed to streamline the integration of AI with enterprise resources by allowing MCP clients to interact with systems like databases and APIs in a secure, structured manner. By using Logic Apps as MCP servers, organizations can reuse existing connectors and workflows, reducing development overhead and accelerating the adoption of AI-powered automation. The feature can be configured by registering Logic Apps connectors as MCP servers through the Azure API Center, which automates much of the setup process.

AWS Launches Kiro: A Spec-Driven AI IDE for Software Development

Amazon Web Services (AWS) has launched Kiro, an AI-powered integrated development environment (IDE) that uses a spec-driven approach to software development. Unlike other AI coding assistants that rely heavily on prompts, Kiro formalizes the development process into three phases: generating user stories, creating a technical design, and breaking the work into a sequence of trackable implementation tasks.

This methodology aims to provide a more structured and agile-friendly workflow for development teams. Kiro is a fork of Visual Studio Code and is designed as a standalone IDE that can be used without an AWS account. The tool is currently in a free preview period, with a free tier and paid plans to follow.

Tigera’s Calico Secures AI Workloads on Kubernetes with Zero-Trust

Tigera has launched a new solution based on its open-source Project Calico software to secure AI workloads running in Kubernetes clusters. The offering addresses the unique security challenges of AI applications, such as unsecured pod-to-pod communication during model training and the protection of sensitive data.

Key features include zero-trust microsegmentation to limit lateral movement by attackers, egress security controls to prevent data exfiltration, and a web application firewall (WAF) to protect AI endpoints from threats like SQL injection. The solution also provides observability and compliance controls through detailed flow logs and visual service graphs, helping teams understand AI service interactions. A cluster mesh capability allows for unified management of security policies across multiple clusters.

Quantum Computing Leaps Forward with Silicon Chip and Supremacy Breakthroughs

Recent developments in quantum computing signal significant progress toward scalable and impactful quantum technologies. Key advancements include a major breakthrough in silicon-based quantum systems by researchers at the University of New South Wales, a demonstration of ‘unconditional’ quantum supremacy by UT Austin and Quantinuum, and a projection by NERSC and Berkeley Lab that practical quantum computing for scientific applications is within a decade’s reach.