AGI, or artificial general intelligence, describes a hypothetical AI system with broad cognitive abilities rather than expertise limited to one task. Such a system could transfer knowledge between unfamiliar problems, learn new skills with limited instruction, reason across domains, and adapt its approach as circumstances change. Today’s language, vision, and robotics models can be highly capable, but they remain uneven and dependent on training patterns, tools, and human-defined objectives. Researchers debate how AGI should be measured, whether human-level performance is the right benchmark, and what technical safeguards would be needed before a generally capable system could be deployed responsibly.
Bristol Myers Squibb and Microsoft have launched a collaboration to accelerate early detection of lung cancer using AI-powered radiology tools, with a focus on underserved communities.
Google released a major upgrade to its Gemini Deep Research agent, giving developers access to advanced autonomous research capabilities through the Interactions API. The company also introduced DeepSearchQA, a benchmark for evaluating multi-step web research agents.
Google has released Gemini 3, its most powerful AI model yet, offering advanced multimodal reasoning, coding capabilities, and agentic planning tools across multiple platforms.
A study by the University of Maryland and Microsoft found that Polish outperforms English and other major languages in long-context AI prompts, surprising researchers.