Societies and the businesses in them benefit from peace and stability. One threat to both is the advent of automated decision environments. Analysts indicate that automated decision environments are intrinsically destabilizing, because they can create unwanted escalation loops. For instance, an automated system may miscalculate whether an incoming warhead is conventional or nuclear. This increases the risk that a conventional conflict accidentally turns into a nuclear one.
To accurately quantify risk, we are developing a monte-carlo simulation environment to predict the effects of decision support automation, cyberattacks, and decision time compression on nuclear strategic stability. Our work is a conceptual successor to simulation models developed by RAND during the cold war. Our simulation tool aims to advise national policy in order to preserve peace and stability.
- an initial literature review of the legal and procedural standards for autonomous weapons.
- the foundation of a simulation tool (concept and software)
- collaborators to help us refine the parameters of our simulation to most accurately match available OSINT about Russian, Chinese, and American developments.
- collaborators who can give guidance on the project in general and potentially join us in authoring the simulation software source code and a paper about it.
If you are interested in collaborating, you can reach us at maarten@snap.com and jon@obsoleteventures.com . We are happy to answer any questions.
Jonathan & Maarten