Artificial Intelligence and Nuclear War Risk 

15 Nov 2023, 5 pm (CET)

by Prof Karl Hans Bläsius

virtual event - ZOOM only

please register at: https://tinyurl.com/akr64ttc   or

https://us02web.zoom.us/meeting/register/tZ0tdu6pqjkiH9DKGzKqedW9JFBG7SPFY0xv

Abstract:

Ever shorter warning times in early warning systems for nuclear threats require increasing use of AI techniques. However, due to the uncertain and incomplete data basis, even AI systems cannot reliably assess threat situations. The risks will continue to increase. This is because an arms race based on the current situation will also occur in connection with AI in particular. This could give rise to weapons systems in the next few years that will be uncontrollable by humans. On May 30, leading AI scientists and AI company executives warned that AI could lead to the extinction of humanity. Developments of specialized non-military AI applications could spiral out of control and dominate the flow of information on the Internet. As a result, social systems could become unstable. This would also significantly increase the risk of nuclear war.

Prof. Dr. Karl Hans Bläsius (Informatics, Hochschule Trier)studied mathematics, physics and computer science at RWTH Aachen University and received a PhD in an area of AI at the University of Kaiserslautern. After industrial activities, he represented "knowledge-based systems" in research and teaching at Trier University of Applied Sciences. He is initiator of unintended-nuclear-war.eu and of quo-vadis-ki.de.

- Intro Slides

- Artificial Intelligence and nuclear war risk (slides , recording )