
Volume 342, Issue 5 · 10 articles from 2 issues
Google unveils Gemini 4 Argon, a frontier model for long-horizon coding, legal, finance and cyber work, rolling out to vetted defenders first. Read more
Google's monthly roundup of September AI launches, led by Gemini 4 Argon, Connected Apps in Gemini and WeatherNext 3. Read more
Google launches Guided Vision in Gemini Live on Android, giving blind and low-vision users real-time conversational visual assistance. Read more
Google DeepMind's SynthID Bio watermarks AI-designed proteins without degrading their biological function. Read more
Gemini adds reusable 'skills' invoked with a slash command, replacing Gems and coming to Google Workspace business customers. Read more
Google rolled out new Connected Apps in Gemini, including Adobe, Airtable, Linear, and Peloton, to manage tasks without switching tabs. Read more
Google Cloud argues scaling OLTP databases for AI agents needs a new architecture that avoids trading off scale, latency, or isolation. Read more
Google Cloud published best practices for using its managed RL fine-tuning service to adapt Gemini models with custom reward functions. Read more
Google pitches its Intelligent Endpoints portfolio, spanning its enterprise browser, ChromeOS, and hardware, as secure ground for agentic AI. Read more
Google shows how its AlphaEvolve evolutionary coding tool can optimize latency-critical video processing beyond what AI coding assistants achieve. Read more