Google DeepMind: A 15-Year Evolution from Atari to EVE Online
Google DeepMind reflects on 15 years of gaming research, showcasing how agents like AlphaZero have paved the way for modern partnerships with game studios.
Google DeepMind Unveils Agentic Video Understanding in Gemini Models
Google DeepMind is transforming video AI by moving from passive frame analysis to agentic, interactive understanding within the Gemini ecosystem.
Google DeepMind Pioneers Double-Blind AI Evaluations to Curb Model Bias
Google DeepMind is revolutionizing AI testing by adopting double-blind evaluation methods, ensuring human assessors remain unbiased when reviewing model performance.
Google Unveils Gemini 3.5 Transcribe: A New Era for AI-Powered Speech-to-Text
Google has officially launched Gemini 3.5 Transcribe, a sophisticated tool designed to revolutionize how we convert spoken language into high-fidelity text.
Google Unveils Gemini 1.5 Flash-8B: A New Era for Scalable AI Development
Google DeepMind has launched Gemini 1.5 Flash-8B, a streamlined, high-efficiency model designed to lower costs and latency for AI developers.
Google’s R4T Breakthrough: Accelerating AI Search Retrieval by up to 20x
Google Research has introduced R4T, a new framework using RL-compiled diffusion to speed up query fan-out by up to 20 times, marking a major leap in AI search efficiency.
Linkup Research Unveils SPARSEUP: A Breakthrough in Sparse Embedding Models
Linkup Research has launched SPARSEUP, an open-source sparse embedding model that delivers industry-leading performance for its size.
Anthropic Launches $5 Million Grant to Study AI’s Impact on Human Wellbeing
Anthropic has announced a $5 million initiative to foster independent research on how AI models influence user wellbeing, aiming to create better industry standards.
WIRED World Fair Heads to Miami: A Convergence of Innovation and Culture
WIRED magazine brings its signature blend of tech, culture, and science to Miami this November for the inaugural WIRED World Fair.
Anthropic Doubles Down on AI Safety: New Initiatives in Alignment and Security
Anthropic is refining its organizational structure to better tackle AI alignment and security, ensuring that rapid innovation remains tethered to human-centric safety standards.
Anthropic Partners with Accenture for Embedded AI Frontier Evaluation
Anthropic is embedding independent evaluators from Accenture directly into its operations to monitor AI safety, training, and development processes.
Anthropic Expands AI Access for Global Research Community
Anthropic is scaling its support for the global scientific community by offering free and discounted Claude subscriptions to researchers worldwide.