Leo Keay and Robert Peters
Limited nuclear war is a possibility grounded in strategic logic and a probability accentuated by the current geopolitical and military context. Planning for limited nuclear war is necessary not only to buttress the credibility of non-strategic nuclear deterrence but also to ensure adequate preparation for a real contingency. The fact that a limited nuclear war has never been fought renders any conclusions about it inevitably provisional. Analytic wargaming and artificial intelligence offer new possibilities for plugging this methodological gap. Tabletop exercises (TTXs) can create an indicative dataset for understanding decision-making dynamics in a hypothetical limited nuclear war. Moreover, computer-enabled large language models trained on the dataset from the TTX could rerun scenarios hundreds of times and adjust the variables to explore alternative outcomes.
No comments:
Post a Comment