Ai

“Agentic AIs are a feral technology… they are projected to become a wild, self-sustaining, population.” Co-organised with Morgan Stanley, this upcoming Santa Fe Institute gathering takes place 15-Sept-2026 in NYC.

A notable June event from the Santa Fe Institute, co-sponsored by OpenAI and Google DeepMind, focused on intelligence across multiple organisational and social scales rather than intelligence as an individual phenomenon.

The paper proposes an “Internet of Agentic AI” in which autonomous agents discover each other, negotiate responsibilities, exchange context, coordinate workflows, and operate across organizational boundaries.

Researchers studying fire-ant collectives found evidence that large-scale ripples and movement waves emerge from simple local alignment between neighbouring ants. No central coordination is required for the collective patterns to appear.

Lantern Pharma announced ZetaSwarm, a commercially available multi-agent scientific reasoning platform designed to run many specialised AI agents in parallel. The system is intended to support drug discovery and biomedical research workflows.

Researchers at the University of Osaka and collaborators have developed a cyborg insect system that reads heartbeat activity, neural signal features, and body movement from cockroaches.

Researchers demonstrated a lightweight navigation system inspired by honeybee learning flights that allows small drones to return home using compact visual memories rather than large-scale mapping systems.

The system uses decentralized coordination and local signaling inspired by ant colony task allocation to distribute warehouse-picking workloads dynamically across autonomous robots.

The advanced capabilities of AI-based navigation in military drones mean they will no longer require human training. They will rely on near-instantaneous sensor data from cameras, LiDAR, and radar to identify obstacles that will inform their future navigation, such as mountains, structures, and other aircraft.

This UK defence programme targets software-defined swarms with both simulation and live-flight trials. The importance is the transition from swarm experimentation toward engineering toolchains and operational validation platforms intended for repeatable deployment and sovereign industrial capability.