Google enhances its Jules AI coding assistant with new command line tools and API support, while Nvidia introduces a revolutionary 4-bit training format that cuts AI training costs, and MIT researchers use generative AI to dramatically accelerate antibiotic discovery from years to much shorter timeframes.
Deloitte deploys Anthropic's Claude AI to 470,000 employees across 150+ countries in the largest enterprise AI rollout ever, creating specialized AI personas for different roles like accountants and developers as the company competes with OpenAI and Google for corporate dominance.
Deloitte Australia offers partial refund for $290,000 government report after academics expose AI-generated fake citations and fabricated quotes falsely attributed to real professors, with the consulting firm quietly admitting to using GPT-4o without disclosure.
OpenAI CEO Sam Altman demonstrates breakthrough AI voice technology speaking fluent Hindi with perfect lip-syncing, while 2024 Turing Award winner Richard Sutton declares large language models a 'dead end' for failing to genuinely learn from experience like humans.
Huawei releases open-source SINQ technology that cuts AI model memory usage by 70%, allowing enterprise models that previously required $30,000 GPUs to run on $1,600 consumer graphics cards, potentially saving thousands in computing costs while maintaining performance quality.
Google AI launches ReasoningBank framework that enables AI agents to self-evolve during operation without retraining by converting past interactions into reusable reasoning strategies, delivering 34% performance improvements and 16% fewer steps across web and software engineering tasks.
Scientists analyzing 8,590 predictions now expect artificial general intelligence to arrive by 2040, dramatically accelerating from previous 2060 estimates, while Anthropic's CEO suggests the singularity could occur within just three months.
Nvidia achieves groundbreaking 4-bit AI training breakthrough with new NVFP4 format, delivering 6X speed boost and 50% memory reduction while matching full precision results on 12-billion-parameter language model, requiring only 1% higher validation loss and 36% less training data than competitors.
Meta releases Code World Model, a groundbreaking 32-billion-parameter AI that learns how code actually executes rather than just syntax, achieving a 65.8% pass rate on coding benchmarks by developing internal world models that enable self-verification and step-by-step simulation.
Major AI systems from Google DeepMind, Microsoft, and Meta generate millions of promising new materials, but scientists discover 80-84% contain fatal flaws that make them impossible to synthesize in reality.