Global AI Weekly

Issue number: 112 | Tuesday, August 26, 2025

Highlights

Meta Freezes AI Hiring as Fear Spreads

Meta Freezes AI Hiring as Fear Spreads

Meta has announced a halt on all hiring within its artificial intelligence division, signaling a cautious approach amid growing concerns in the tech industry. This freeze highlights the challenges and uncertainties surrounding AI development and its broader implications. The move reflects Meta's strategic reevaluation as companies grapple with balancing innovation and responsibility in this rapidly evolving space.

futurism.com

OpenAI introduces new ChatGPT features and brings back GPT-4o

OpenAI introduces new ChatGPT features and brings back GPT-4o

OpenAI is enhancing the personality of GPT-5, aiming to make it feel more approachable and warmer for users. In addition, the company is reintroducing GPT-4o alongside new features for ChatGPT, focusing on improving user experience and interaction. These updates are part of OpenAI's ongoing effort to refine and expand its AI offerings.

techzine.eu

Research

MCP for Research: How to Connect AI to Research Tools

MCP for Research: How to Connect AI to Research Tools

Explore how the Modular Control Panel (MCP) connects artificial intelligence to research tools, making it easier for researchers to integrate AI into their workflows. This approach supports collaboration and innovation by leveraging open source and open science principles. It’s all about simplifying complex processes and empowering the research community with accessible AI technology.

huggingface.co

Speed Always Wins: A Survey on Efficient Architectures for Large  Language Models

Speed Always Wins: A Survey on Efficient Architectures for Large Language Models

This paper explores the importance of speed in designing efficient architectures for large language models. It highlights various strategies and techniques to enhance performance while maintaining accuracy and scalability. The discussion focuses on balancing computational efficiency with the growing demand for powerful AI systems.

huggingface.co

Video

AI Show: Trustworthy Agents

AI Show: Trustworthy Agents

In this episode of The AI Show, Sarah Bird discusses Trustworthy Agents, focusing on the importance of building secure AI systems while addressing risks like misuse, malfunctions, and systemic issues. She highlights strategies for mitigating these challenges and ensuring reliability through trust, security, and adherence to regulations. The episode offers insights into creating responsible AI solutions and effectively managing their lifecycles.

youtube.com

Taking Agents everywhere with Semantic Kernel

Taking Agents everywhere with Semantic Kernel

Microsoft Senior Software Engineer Eduard van Valkenburg presents "Taking Agents Everywhere with Semantic Kernel" at AgentCon, showcasing the architecture and versatility of Semantic Kernel. This talk explores creating plugins, setting up AI agents, and facilitating multi-agent interactions, with support for programming languages like Python, .NET, and Java. It's a practical session for developers aiming to connect API services or design custom agent functions using this modular tool. AgentCon brings together global developers and researchers to explore cutting-edge advancements in autonomous AI systems.

youtube.com

Articles

Hugging Face Unveils AI Sheets: A Free, Open-Source No-Code Toolkit for LLM-Powered Datasets

Hugging Face Unveils AI Sheets: A Free, Open-Source No-Code Toolkit for LLM-Powered Datasets

Hugging Face introduces AI Sheets, a no-code, open-source tool designed to make data exploration and manipulation easier for everyone. Combining the familiarity of spreadsheets with the power of large language models (LLMs), it allows users to analyze, clean, and augment datasets without requiring advanced coding skills. This free toolkit is designed to make AI-driven data workflows more accessible to a broader audience.

marktechpost.com

Alibaba AI Team Just Released Ovis 2.5 Multimodal LLMs

Alibaba AI Team Just Released Ovis 2.5 Multimodal LLMs

Alibaba's AI team has introduced Ovis 2.5, a cutting-edge multimodal large language model that pushes the boundaries of open-source AI. This release brings upgraded visual perception and reasoning abilities, making it more powerful and versatile for a range of applications. With these enhancements, Ovis 2.5 aims to contribute significantly to advancing AI accessibility and functionality.

marktechpost.com

GPT-5: The Reverse DeepSeek Moment

GPT-5: The Reverse DeepSeek Moment

The release of GPT-5 has sparked widespread consensus about its mishandling, particularly with the abrupt leap from GPT-4o and GPT-3. The article reflects on how this transition lacked the same impactful strides seen in prior iterations, raising questions about its development and rollout. It offers a thought-provoking perspective on the expectations and challenges faced in advancing AI models.

lesswrong.com

Upcoming Events

AgentCon 2025

AgentCon 2025

After an exciting set of AgentCon events before the summer holidays, AgentCon is happening in more locations during the second half of the year. The AI Agents World Tour is a series of one-day conferences for developers focused on autonomous AI agents. Spanning cities like San Francisco and Singapore, these events gather top engineers, researchers, and creators shaping the future of AI technologies. It’s an opportunity to connect, learn, and innovate in the fast-evolving world of autonomous AI. Check all events and register for an event near you!

globalai.community

Code

Model Context Protocol (MCP) is Now Generally Available in Visual Studio

Model Context Protocol (MCP) is Now Generally Available in Visual Studio

Model Context Protocol (MCP) is now generally available in Visual Studio, bringing enhanced real-time insights across your development stack. This update supercharges the capabilities of Agent Mode, providing developers with richer context and streamlined workflows. It’s a major step forward in making development smarter and more efficient!

devblogs.microsoft.com

GPT-OSS - A C# Guide with Ollama

GPT-OSS - A C# Guide with Ollama

Discover how to run GPT-OSS locally using C# and Ollama in this informative guide. The setup enables fast, private, and offline AI capabilities, making it ideal for projects that require confidentiality and speed. Learn the essential steps to integrate GPT-OSS into your .NET applications and explore the potential of offline AI in a secure environment.

devblogs.microsoft.com

Podcast

Google DeepMind: The Podcast

Google DeepMind: The Podcast

What is generative AI? How do you create safe and capable models? Where is the industry headed in the next decade? Join Professor Hannah Fry as she uncovers the extraordinary way AI is transforming our world.

deepmind.google

>