News

the-decoder3d
Meta releases first multimodal Llama-4 models, leaves EU out in the cold
Meta has released the first two models in its Llama 4 series, marking the company’s initial deployment of a multimodal architecture built from the ground up. Both models—Llama 4 Scout and Llama 4 ...
the-decoder3d
Microsoft releases real-time AI-generated playable demo of Quake II
Microsoft has introduced a research project that generates and runs Quake II entirely within an AI model, producing a playable version of the game in real time. The model, called WHAMM (World and ...
the-decoder4d
Anthropic study finds language models often hide their reasoning process
A new Anthropic study suggests language models frequently obscure their actual decision-making process, even when they appear to explain their thinking step by step through chain-of-thought reasoning.
the-decoder5d
OpenAI plans GPT-5 release in "a few months," shifts strategy on reasoning models
In a post on X, CEO Sam Altman announced that the company will release its reasoning models o3 and o4-mini as standalone systems in the coming weeks—reversing earlier plans to integrate them directly ...
the-decoder3d
Google adds web search capabilities to NotebookLM
Google has expanded NotebookLM's functionality with a new web search feature that helps users discover and incorporate online sources directly into their notebooks. The system lets users describe what ...
the-decoder4d
Meta reportedly prepares to launch Llama 4 after delays and internal overhaul
Meta’s next language model, Llama 4, is reportedly nearing release after repeated delays and significant changes to its development process. According to The Information, the launch is now planned for ...
the-decoder4d
Google expands access to Gemini 2.5 Pro amid strong benchmark results
According to Alphabet CEO Sundar Pichai, Gemini 2.5 Pro represents Google's "most intelligent model + now our most in demand." Demand has increased by over 80 percent this month alone across both ...
the-decoder6d
Microsoft reportedly pulls back on global data center expansion
Microsoft has halted or delayed multiple data center projects worldwide, raising questions about the current state of AI service demand. According to Bloomberg, Microsoft is stepping away from or ...
the-decoder5d
Midjourney launches v7 image model with draft mode for faster iteration
According to the company, the new version is designed to improve both text and image prompt handling, with more coherent and detailed renderings of textures, bodies, hands, and other objects. For the ...
the-decoder5d
Microsoft adds ChatGPT-style features to expand Copilot's capabilities
Microsoft is expanding its Copilot assistant with a range of new features, many of which resemble functions already available in ChatGPT. Others, however, are new additions that aim to personalize and ...
the-decoder6d
LLMs struggle to match human researchers in paper replication test
OpenAI's new PaperBench benchmark reveals the current limitations of AI's ability to independently replicate scientific research, with human researchers still maintaining an edge. The benchmark puts ...