PRIME-RL launches as a fully open-source reinforcement learning framework capable of scaling to 1000+ GPUs and trillion-parameter models, supporting major AI model families with asynchronous training, multimodal capabilities, and flexible multi-node deployment under the Apache 2.0 license.
Meta taps 28-year-old Scale AI co-founder Alexandr Wang to lead a secretive AI lab called TBD, assembling 100 elite researchers and releasing a competitive new model as Zuckerberg races to close the gap with OpenAI and Google — but internal tensions, setbacks, and rival speed are threatening to slow the …
Chinese AI startup DeepSeek is raising $7.4 billion in its first-ever funding round, valuing the company at up to $59 billion, with founder Liang Wenfeng personally committing $3 billion and Tencent eyeing a major stake, as China accelerates its push for AI dominance following DeepSeek's models shaking confidence in U.S. …
Google DeepMind's Logan Kilpatrick, a veteran of NASA, OpenAI, and Apple, is urging developers to think bigger as AI tools like Gemini 2.5 Flash and Project Antigravity redefine software engineering, drawing a sharp line between casual 'vibe coding' and serious agentic development.
Google DeepMind launches Gemma 4 12B, a multimodal AI model that runs locally on consumer laptops with just 16GB of VRAM, featuring a groundbreaking encoder-free architecture that processes vision and audio directly through its LLM backbone for lower latency and memory use, available now on Hugging Face and Kaggle under …
Ideogram 4 launches as a 9.3B parameter open-weight text-to-image model that outperforms rivals up to 80B parameters, offering 2K resolution, multilingual text rendering, and spatial layout control — now freely available on GitHub and Hugging Face.
A new attention variant called Wall Attention launches with per-channel, per-timestep multiplicative decay for independent forgetting rates, backed by optimized Triton kernels enabling efficient autoregressive decoding, full GQA support, attention sinks, sliding windows, and sequence packing with verified numerical accuracy.
Microsoft unveils a massive AI overhaul at its Build conference, introducing personal agents, a 35-billion-parameter reasoning model, and open-source security tools as CEO Satya Nadella positions the tech giant as a dominant force in the rapidly evolving enterprise AI market.
Microsoft Build 2026 unveils a sweeping AI expansion, launching new MAI reasoning models, GitHub Copilot desktop tools, a full-stack developer platform, and agentic scientific research systems now powering real-world breakthroughs at organizations like GSK and BHP.
NVIDIA launches Cosmos 3, the world's first fully open omnimodel capable of natively processing text, images, video, sound, and actions for robotics and autonomous vehicles, while simultaneously unveiling the global Cosmos Coalition to accelerate open world model development across industries.