48 articles
Google DeepMind has introduced a new framework for testing artificial general intelligence (AGI) that functions like an IQ test for AI. It evaluates systems across 10 cognitive abilities, comparing them to human performance to create a detailed profile rather than relying on single scores. This aims to bring scientific rigor to the debate on AI progress.
Hermes Agent, developed by News Research, introduces persistent memory and autonomous skill development, allowing AI to continuously improve. It offers flexible model choices, including free options, and is designed for easy setup on local machines or VPS.
New AI models can now learn and adapt like humans, building on past knowledge without forgetting. This "continual learning" avoids the costly retraining needed for current AI and promises more versatile and efficient AI systems for various applications.
New artificial intelligence models are demonstrating remarkable progress in understanding complex instructions and generating human-like text. These advanced Large Language Models (LLMs) are setting new benchmarks and becoming more accessible for individuals and businesses. This leap forward promises to enhance everyday tasks, automate complex processes, and drive innovation across various industries.
Google has released Gemma 4, a powerful and truly open-source AI model designed to run on consumer hardware. This tiny model challenges existing AI norms by offering high performance with minimal resource requirements, making advanced AI more accessible to developers and users.
Google's new TurboQuant technique promises to drastically cut the memory needed for AI models, potentially making them cheaper and faster to run. Early tests show significant memory savings and speed improvements, though the full impact is still being evaluated.
Google's new Turboquant algorithm drastically cuts the memory needed for AI, potentially solving a major bottleneck. This breakthrough promises longer AI conversations, AI on personal devices, and lower costs, impacting the semiconductor industry.
Google's new TurboQuant technology dramatically reduces AI memory needs and boosts processing speed by up to 8x without losing accuracy. This breakthrough promises significant cost savings and expanded capabilities for AI applications.
A new AI tool called AutoResearch, now part of the AgentZero project, allows artificial intelligence models to learn directly from data scraped from the internet. This capability enables AI to access and process real-time information, significantly enhancing its accuracy and relevance for complex tasks and current events.
Anthropic's most powerful AI model, Claude Mythos, has reportedly been leaked online. This new model shows significant advancements in coding, reasoning, and cybersecurity, surpassing previous versions like Claude Opus 4.6. Anthropic is proceeding with caution due to potential risks, especially in cybersecurity, and is working on efficiency and cost before a wider release.
