Robophilosophy 2026 podcast

3. John P. Sullins and Colette Cadeaux

0:00
25:59
Spol 15 sekunder tilbage
Spol 15 sekunder frem

Project STORM is a prototype risk‑management framework to support responsible robotics in high‑stakes national security contexts. John and Colette report on some of the outcomes of this project with a focus on security and defence AI/robotics applications. They highlight certain risks such as amplified bias, data poisoning, and inadequate testing that can lead to system failures or misidentification of targets in military AI/robotics systems. They find that traditional “human‑in‑the‑loop” safeguards are declining in defence applications, with the industry shifting toward human–machine teaming or “centauring”. This move has a new set of ethical issues, which they assess using STORM’s prototype risk management framework.

Flere episoder fra "Robophilosophy 2026"