The LLM archive monitors model releases, quantization updates, reasoning capabilities, and real-world deployment implications for local and hybrid AI. We focus on what materially changes selection and operations: context windows, latency, memory footprint, licensing, and evaluation evidence across open and commercial families. This section is designed for teams that need dependable model intelligence, not hype cycles. Pair these updates with the LLM pillar and references to hardware constraints and framework integration.
As AI, cloud, and other technology investments soar, organizations have to make investment decisions with increased speed and clarity. Practices like FinOps, IT financial management (ITFM), and strategic portfolio management (SPM) help stakeholders evaluate opportunities and trade-offs for maximum value. But they depend on unified, reliable data.
OpenAGI has announced the release of Lux, an AI model that claims to have a 83.6% success rate in executing commands on a computer, surpassing models by OpenAI and Anthropic.
Liquid AI, a MIT spinoff, has introduced small-model training for enterprises
DeepSeek, a Chinese startup, has released a top-tier AI model that rivals those of leading American companies without exclusive costs. The model, called DeepSeek-V3.2, uses open-source technologies and achieves comparable performance to premium commercial models.
OpenAI is awarding up to $2 million in grants for research at the intersection of AI and mental health. The program supports projects that study real-world risks, benefits, and applications to improve safety and well-being.
This article describes the development of Primus-Turbo, an open-source library for large-scale pre-training with Primus-Turbo, and its use with AMD Instinct MI325X. The goal is to optimize the performance of MoE pre-training.
Accenture and OpenAI collaborate to help enterprises bring agent AI capabilities into the core of their business and unlock new levels of growth.
OpenAI and NORAD are bringing new magic to “NORAD Tracks Santa” with three ChatGPT holiday tools that let families create festive elves, toy coloring pages, and custom Christmas stories.
US telecom company uses AI to scan inmate calls for planned crimes
Digital twins are transforming manufacturing with real-time simulation and optimization capabilities.
Digital resilience—the ability to prevent, withstand, and recover from digital disruptions—has long been a strategic priority for enterprises. With the rise of agentic AI, the urgency for robust resilience is greater than ever.
The AlphaFold project won the chemistry Nobel and revolutionized biology. But what's next? The project's founder, John Jumper, talks about his expectations for the future.
This investigation explores the implications of privacy for chatbot companions, which are revolutionizing our daily lives. But at what cost? Can we still protect our privacy in a world increasingly integrated with AI?
L'indice di ipotesi sull'intelligenza artificiale è stato creato per fornire un riassunto rapido e chiaro dell'industria. Tuttavia, l'industria sta facendo strani passi con la creazione di contenuti automaticamente generati.
Le aziende investono miliardi di dollari in agenti AI e infrastrutture per trasformare i processi aziendali, ma si riscontrano limitate successi nella realtà, a causa dell'impossibilità degli agenti di capire veramente i dati aziendali, le politiche e i processi. L'ontologia è la chiave per evitare che gli agenti AI sbagliano.
Andrej Karpathy, Tesla's former AI director and OpenAI co-founder, wrote a code behind which a 'vibecode' for creating an AI orchestration system. The project explores the role of AI model management in corporate and highlights the need for governance in the industry.
Researchers at the University of Science and Technology in China developed a new reinforcement learning framework that helps train large language models (LLMs) for complex agentic tasks beyond well-defined problems like math and coding. The framework, called Agent-R1, is compatible with popular RL algorithms and shows significant improvements on reasoning tasks that require multiple retrieval stages and multi-turn interactions with tools.
Summary in English of the 2025 year: a year of technological progress and diversity in the AI world.
Anthropic ha sviluppato un nuovo framework per gestire la memoria degli agenti AI, risolvendo il problema della loro capacità di ricordare istruzioni e conversazioni durante sessioni lunghe. Il nuovo SDK è stato testato con successo su diverse applicazioni.
Observable AI is essential for making large language models (LLMs) reliable and trustworthy. This article explores how to apply observability principles to ensure security, transparency, and accountability in AI decision-making processes.