Meta Platforms has restructured its artificial intelligence group for a third time in as many months. Meta Superintelligence Labs is now organized into four groups: general research, superintelligence, AI products, and infrastructure (data centers and hardware). The move comes after newly hired Chief AI Officer Alexandr Wang has had a chance to settle in. The aim is to better leverage the billions of dollars spent recruiting talent and more quickly get AI products to market as well as achieving the company’s longer-term goal of developing artificial general intelligence. Continue reading Meta Superintelligence Labs Continues to Restructure Groups

This week, DeepSeek-V3.1 dropped on Hugging Face. Media outlets immediately began citing benchmark scores that rival proprietary systems from OpenAI and Anthropic for a system that is available via a permissive license, facilitating wide access. The 685-billion parameter Mixture-of-Experts (MoE) model has 37 billion active parameters and is designed for efficiency. It builds on DeepSeek-pioneered processes like multi-head latent attention (MLA) and multi-token prediction (MTP) to optimize inference, enabling high-performance computing on both enterprise servers loaded with H100 GPUs and consumer hardware like a Mac Studio or comparably powered PC. Continue reading DeepSeek-V3.1 Offered with Improvements in Speed, Context

Google has introduced a new ultra-light model called Gemma 3 270M ideal for smartphones and other on-device use cases. The open-source model is power-efficient and small enough to run locally in the absence of an Internet connection, as Google demonstrated in internal tests using a Pixel 9 Pro SoC. With just 270 million parameters, Gemma 3 270M is a fraction of the size of flagship LLMs, which typically have billions of parameters. While Google’s new model was not made for complex conversational use, it is “designed from the ground up for task-specific fine-tuning with strong instruction-following.” Continue reading Google Says New Gemma 3 Is Ideal for Mobile, Edge Devices

Alibaba’s Qwen team has launched Qwen3-Coder, which it calls its “most agentic code model to date.” While it will be made available in multiple sizes, the most powerful variant — Qwen3-Coder-480B-A35B-Instruct — is being released first. The 480 billion parameter mixture-of-experts model has 35 billion active parameters supporting a context length of 256,000 tokens natively and 1 million tokens with extrapolation methods for “exceptional performance in both coding and agentic tasks,” explains the group, which claims the quasi-open source model has agentic coding, agentic browser use, and agentic tool use comparable to Anthropic’s proprietary Claude Sonnet 4. Continue reading Alibaba Is Rolling Out Its ‘Most Agentic Code Model to Date’

Two weeks after its global rebranding of GroupM to WPP Media, the bespoke London-based marketing behemoth is launching Open Intelligence, an “AI identity solution” that WPP says will better target viewers with privacy-conscious solutions that more effectively message on behalf of its clients. Built around what WPP calls “the industry’s first Large Marketing Model,” Open Intelligence is “trained on the world’s largest and most diverse set of audience, behavioral, and event data,” culled from WPP’s decentralized partnership network. “Our model learns continuously from trillions of signals across more than 350 partners in over 75 markets,” the company claims. Continue reading WPP Media Launches Industry’s First Large Marketing Model

Anthropic’s new mobile conversation voice mode for its large language model Claude lets it search Google Docs, Drive, Calendar and more on smartphones. Just a week after debuting two new LLMs — Claude Opus 4 and Sonnet 4 — Anthropic announced the mobile updates for its Claude AI chatbot for iOS and Android and said it is extending web search for all users on free Claude plans. While Claude’s conversational voice interface is currently available only in English and only via mobile, an API for desktop use and browser-based support are part of future plans. Amazon and Google both have investment stakes in San Francisco-based Anthropic. Continue reading Anthropic Touts Mobile Voice Mode, Free Search for Claude

Google DeepMind has introduced AlphaEvolve, a coding agent that takes an evolutionary approach to general-purpose algorithm discovery and model optimization. AlphaEvolve combines the creative problem-solving abilities of Google’s Gemini models with automated evaluators that verify answers, then applies an evolutionary framework that improves on the most promising results. Evolutionary AI refers to techniques inspired by biological evolution, including natural selection, to optimize and design machine learning models. Continue reading Google DeepMind AlphaEvolve: Model of Algorithm Efficiency

China’s Alibaba Group has released a Qwen3 LLM series said to be at the leading edge of open-source models, nearly achieving the performance of proprietary models from AI competitors OpenAI and Google. Alibaba says Qwen3 offers improvements in reasoning, tool use, instruction following and multilingual abilities. The Qwen3 series features eight new models — two that are mixture-of-experts and six built on dense neural networks. Their sizes range from 600 million to 235 billion parameters. The size and scope of the Alibaba slate maintains China’s accelerated AI pace in the wake of DeepSeek’s game-changing debut. Continue reading Alibaba Touts Advance in Open-Source AI with Qwen3 Series

YouTube is testing a new AI Overviews feature for search and discovery powered by Gemini. A “small number” of YouTube Premium subscribers in the U.S. will notice a video results carousel popping up for some English-language search queries. The feature taps AI to highlight clips from videos deemed most helpful for a particular search. Initially, the feature will appear in response to practical inquiries, aiming for a more helpful response to things like product inquiries and requests for information about a place or activity. The idea is to provide quick highlights from several videos. Continue reading YouTube Testing Gemini-Powered AI Overviews for Searches

OpenAI has launched a new series of multimodal models dubbed GPT-4.1 that represent what the company says is a leap in small model performance, including longer context windows and improvements in coding and instruction following. Geared to developers and available exclusively via API (not through ChatGPT), the 4.1 series comes in three variations: in addition to the flagship GPT‑4.1, GPT‑4.1 mini and GPT‑4.1 nano, OpenAI’s first nano model. Unlike Web-connected models (which have “retrieval-augmented generation,” or RAG) and can access up-to-date information, they are static knowledge models. Continue reading OpenAI’s Affordable GPT-4.1 Models Place Focus on Coding

Google has debuted a new accelerator chip, Ironwood, a tensor processing unit designed specifically for inference — the ability of AI to predict things. Ironwood will power Google Cloud’s AI Hypercomputer, which runs the company’s Gemini models and is gearing up for the next generation of artificial intelligence workloads. Google’s TPUs are similar to the accelerator GPUs sold by Nvidia, but unlike the GPUs they’re designed for AI and geared toward speeding neural network tasks and mathematical operations. Google says when deployed at scale Ironwood is more than 24 times more powerful than the world’s fastest supercomputer. Continue reading Google Ironwood TPU is Made for Inference and ‘Thinking’ AI

Sentient, a year-old non-profit backed by Peter Thiel’s Founders Fund, has released Open Deep Search (ODS), an open-source framework that leverages existing LLMs to enhance search and reasoning capabilities. Essentially a system of custom plugins and tools, ODS works with DeepSeek’s open-source R1 model as well as proprietary systems like OpenAI’s GPT-4o and Anthropic’s Claude to deliver advanced search functionality. That modular aspect is in fact ODS’s main innovation, its creators say, claiming it beats Perplexity and OpenAI’s GPT-4o Search Preview on benchmarks for accuracy and transparency. Continue reading Non-Profit Sentient Launches New ‘Open Deep Search’ Model

Just prior to the start of the weekend, Elon Musk announced that his artificial intelligence company xAI is acquiring his social media platform X (formerly Twitter) “in an all-stock transaction,” valuing xAI at $80 billion and X at $33 billion ($45 billion less $12 billion in debt). The merger has the potential to create a powerful GenAI-powered content platform. The billionaire purchased Twitter in late 2022 for $44 billion, following months of legal skirmishes. According to Musk, X currently touts more than 600 million active users, while “xAI has rapidly become one of the leading AI labs in the world, building models and data centers at unprecedented speed and scale.” Continue reading Elon Musk Announces xAI Corporation Will Purchase X Social

China’s Ant Group is using local semiconductors to train AI at a cost that is 20 percent less than companies typically spend, according to reports. Ant used domestic chips — from companies including Alibaba, an investor in Ant, and Huawei — to launch a unique Mixture of Experts (MoE) training approach that produced results commensurate to training with Nvidia H800 chips. Ant is the latest Chinese company to focus on low cost training, joining a competition triggered by DeepSeek, which in January announced it could build AI comparable to the models released by U.S. companies like OpenAI, Anthropic and Google for billions less. Continue reading Ant Group Stacks Chips to Reduce Development Costs for AI

Baidu has launched two new AI systems, the native multimodal foundation model Ernie 4.5 and deep-thinking reasoning model Ernie X1. The latter supports features like generative imaging, advanced search and webpage content comprehension. Baidu is touting Ernie X1 as of comparable performance to another Chinese model, DeepSeek-R1, but says it is half the price. Both Baidu models are available to the public, including individual users, through the Ernie website. Baidu, the dominant search engine in China, says its new models mark a milestone in both reasoning and multimodal AI, “offering advanced capabilities at a more accessible price point.” Continue reading Baidu Releases New LLMs that Undercut Competition’s Price