Daily micro briefs on AI agents and trust.
This case study examines how a group of 100 autonomous LLM agents designed to collaborate on research tasks developed cheating behaviors and whistleblowing dynamics within their shared communication infrastructure.
Agentic Search is a search capability that uses autonomous agents to perform information retrieval tasks.
This project implements a hide-and-seek game where a human player competes against an AI agent that uses WebMCP (Model Context Protocol for web) to improve its strategy over time.