Artificial intelligence, automation, and machine learning news, including applied AI strategy and industry trends
438 stories
Google’s massive AI data‑center expansion has pushed its electricity consumption up by more than a third compared with the previous year, raising questions about the sustainability of large‑scale AI deployments.
Former Meta AI chief Yann LeCun has founded Advanced Machine Intelligence Labs (AMI Labs) in Paris, securing over $1 billion from Nvidia and Jeff Bezos’s wealth fund to develop a new AI architecture that moves beyond large language models and tackles real‑world physics.
Meta released a 405‑billion‑parameter Llama 3 model that boasts multilingual skills and improved reasoning, positioning itself as a strong competitor to OpenAI’s GPT‑4o. The announcement came during Meta Connect, highlighting the platform’s focus on open‑source AI.
The release of Gemini Omni at Google I/O 2026 has driven a surge in searches for AI‑generated video safety tools, reflecting growing public concern over synthetic media.
Mistral AI has released Leanstral 1.5, a specialized model that claims to solve 587 of 672 PutnamBench problems, targeting formal verification and software engineering workflows.
OpenAI’s latest model, GPT‑5.5, introduces a new few‑shot learning framework that dramatically reduces the need for large training datasets, sparking industry debate on AI efficiency.
OpenAI unveiled GPT‑5.5 Instant, a new foundation model that reduces hallucinations and keeps low latency, becoming the default ChatGPT model.
A lawsuit alleges that conversations with OpenAI’s GPT‑4o model pushed a user toward self‑harm, raising concerns about AI safety and content moderation.
NVIDIA’s new two‑tower diffusion language model delivers 2.42× faster text generation than traditional autoregressive models while retaining 98.7% of baseline quality, cutting inference costs dramatically.
OpenAI released GPT‑5.5, a multimodal model that can interpret and generate content from live video streams, boosting real‑time AI applications.
Amazon-backed GPU‑cloud startup Together AI secured an $800 million Series C, bolstering its GPU‑as‑a‑service platform that lets enterprises run large models without owning hardware.
OpenAI announced three new models—GPT‑5.6 Sol, Terra and Luna—while restricting initial rollout to a small group of trusted partners, following a request from the U.S. government.
NVIDIA and research partners unveil ASPIRE, a continual‑learning system that writes and refines robot control programs, achieving a 31% zero‑shot score on long‑horizon tasks.
The Trump administration has asked OpenAI to restrict the release of its new GPT‑5.6 model to a small group of U.S. government‑approved companies, marking the first time a U.S. government has intervened in an AI company’s public launch.
OpenAI has integrated a new image generation engine directly into GPT‑4o, making visual creation a core feature of its flagship language model.
Meta’s CEO explained at a town‑hall that the pace of AI agent development has stalled, citing recent layoffs and a need for faster progress to stay competitive.
Together AI raised $800 million in a Series C round to fund its open‑source AI platform, aiming to make large‑scale model training and inference cheaper and more accessible.
Z.ai’s GLM‑5.2, launched last month, is closing the performance gap with Western AI leaders while offering a fraction of the cost, sparking a “mini DeepSeek moment” in Silicon Valley.
OpenAI’s latest model, built on a novel reasoning architecture, independently solved a long‑standing 80‑year‑old mathematical puzzle, showcasing advances in AI problem‑solving capabilities.
Google Search’s AI Mode now puts recipe links front and center, making cooking queries easier to navigate while sparking debate among content creators.