AI agents achieve 100% success rates on complex investigatory tasks despite only 60% accuracy in data extraction by routing around failures and self-correcting, proving that hallucinations matter less than expected in observability systems where data is already compressed and lossy.
Groundbreaking research reveals that AI language models wastefully activate their entire networks for social reasoning tasks while actually using only a tiny specialized subset of parameters, creating massive inefficiency compared to human brains and pointing toward revolutionary energy-saving AI designs.
Scientists develop world's first microwave-powered computer chip that processes over 20 billion operations per second using analog microwaves instead of digital circuits, consuming less than 200 milliwatts while achieving 88% accuracy in pattern recognition tasks for AI and wearable applications.
New research testing 24 advanced AI language models reveals they systematically fail to distinguish between belief and knowledge, with accuracy plummeting from over 90% to as low as 14.4% when processing first-person false beliefs, exposing critical flaws in their reasoning capabilities.
Collins Dictionary crowns 'vibe coding' as Word of the Year, recognizing the revolutionary practice of creating computer programs by simply describing them to AI instead of manually writing code, making app development accessible to non-programmers.
WisdomAI secures $50 million Series A funding led by Kleiner Perkins and NVentures as the startup explodes from 2 to 40 enterprise customers including Cisco and ConocoPhillips with its hallucination-free AI analytics platform that lets users query business data in natural language.
MIT researchers unveil SEAL, a groundbreaking technique that enables AI models to permanently learn new information by creating their own study sheets and updating internal weights, boosting accuracy by 15% on question-answering and over 50% on skill-learning tasks.
Chinese AI firm Moonshot AI releases Kimi K2 Thinking, an open-source reasoning model that outperforms GPT-5 and Claude Sonnet 4.5, achieving record 44.9% on Humanity's Last Exam benchmark and marking a potential turning point as Chinese models now rival leading US systems.
GitHub unveils AgentHQ at GitHub Universe 2025, a new platform that enables developers to create AI agents capable of automating complex development tasks like issue triage, code reviews, and deployment beyond what Copilot offers, marking a significant leap toward autonomous software development workflows.
AI adoption skyrockets to 88% of organizations as companies rush to experiment with artificial intelligence, but only one-third successfully scale implementations enterprise-wide, with most struggling to move beyond pilot phases despite heavy investment in the transformative technology.