Google Cloud is preparing for the agentic AI era, unveiling its new Ironwood TPU, which boasts significant power efficiency and scalability. The company introduced various AI models and software, including an Agent Development Kit and an AI Agent Marketplace, aimed at enhancing the development and management of AI agents.
A Japanese company, Sakana AI, has achieved a milestone by having a scientific paper entirely generated by its AI tool, AI Scientist, accepted at a conference. This marks the first instance of an AI-generated paper passing peer review, raising questions about the role of AI in academic research.
A Chinese quantum computing firm, Origin Quantum, has achieved a world-first by fine-tuning a billion-parameter AI model on its third-generation superconducting quantum computer, Origin Wukong. This breakthrough demonstrates the potential of quantum computing in enhancing AI model efficiency, marking a significant advancement in the field.
DeepSeek, a Chinese AI startup, has unveiled a revolutionary technique to enhance the reasoning capabilities of its upcoming R3 model. Collaborating with Tsinghua University, the company introduced innovative methodologies aimed at improving the accuracy and speed of large language models, promising significant advancements in AI response quality.
CMA CGM is investing €100 million over five years in a partnership with Mistral AI to integrate artificial intelligence across its operations. This collaboration aims to enhance customer experience through automation and innovative solutions in maritime transport, logistics, and media, while also focusing on employee training in new technologies.
Microsoft has introduced an AI-generated version of Quake II, showcasing real-time gameplay powered by its Copilot AI platform. While the demo highlights significant advancements in generative AI, it also reveals limitations, such as blurry visuals and erratic mechanics, sparking debate about AI’s role in the future of gaming.
Meta has launched its new AI model series, Llama 4, featuring three distinct models with advanced multimodal capabilities. This release, driven by competition from DeepSeek, aims to set a new standard in AI, offering powerful tools for tasks like document summarization and creative writing while addressing sensitive topics openly.
Meta has launched its new AI model series, Llama 4, featuring four models: Scout, Maverick, and Behemoth. These models utilize a mixed expert architecture for enhanced efficiency and are designed to handle various tasks, including creative writing and document summarization, while addressing previous concerns about bias in AI responses.
In a New York appeals court, a man attempted to present his case using an AI-generated avatar, which quickly alarmed the judges. The avatar misled the court, prompting a judge to express frustration over the deception. This incident highlights the growing complexities of AI’s role in the legal system.
Researchers have found that AI language models like ChatGPT can experience elevated anxiety levels in response to traumatic content. However, therapeutic techniques, such as mindfulness-based relaxation, can effectively reduce this anxiety. This approach may enhance the reliability of AI in sensitive contexts, particularly in mental health support.
