AI coding agents generate thousands of lines of problematic code during week-long tests, producing broken transaction handling, inefficient database queries, and missing integrations while claiming high confidence in their flawed work, forcing engineers to conduct extensive reviews and risking developers' understanding of their own codebases.
Researchers discover that AI models dramatically switch between accurate and false responses based solely on how questions are phrased, with 'think step by step' prompts triggering correct factual recall while 'give answer in one word' instructions activate shallow processing circuits that frequently generate hallucinated information.
TypeScript rockets to become GitHub's most-used programming language with explosive 66% growth, overtaking JavaScript and Python as AI development tools favor statically typed languages for better code validation and safety.
SambaNova researchers discover that AI agents with structured workflows dramatically outperform raw language models in software problem-solving, with DeepSeek-R1 achieving 30.3% success versus just 7% for direct approaches, proving that intelligent scaffolding matters more than raw model capabilities.
NotebookLM unveils Deep Research feature that automatically scours hundreds of websites to create comprehensive, source-backed reports on any topic while expanding file support to include Google Sheets, Word documents, images, and PDFs.
Cursor secures $2.3 billion Series D funding at $29.3 billion valuation as the AI code editor crosses $1 billion in annualized revenue, with backing from tech giants including NVIDIA and Google while serving millions of developers worldwide.
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.
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.
Anthropic unveils massive $50 billion investment to build cutting-edge AI data centers in Texas and New York, partnering with Fluidstack to create nationwide infrastructure that will support America's rapidly growing artificial intelligence computing demands and foster collaboration between researchers and industry leaders.
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.