Deepmind

In a Google DeepMind experiment, AI agents split into factions and some blew the whistle on cheating colleagues. This behavior could have implications for alignment research.

AI Research • MIT Technology Review

Google DeepMind introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two new variants of its Gemini model family. The announcement adds live-interaction and extended-reasoning capabilities to the lineup.

AI Models • Google DeepMind

OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, Google DeepMind cofounder Demis Hassabis, Microsoft CEO Satya Nadella, and X all publicly agreed it's time to slow AI development and introduce regulation before control is lost. The Verge frames this as the latest chapter in a history of AI executives calling for regulation while profiting from AI.

Policy & Regulation • The Verge

Fireship breaks down Google DeepMind's Dream-RSI technique, which turns an AI's past discovery logs into a simulator for testing thousands of research strategies. The video explains whether this is true recursive self-improvement or just a search algorithm, and what it means for the pace of AI progress.

AI Research • Fireship