I disagree. While I concede that, if there is credible information that an adversary is developing AI weapons that need to be counterbalanced to reach some sort of Nash equilibrium, simply putting time and effort into the development of any adversarial technology will turn up the temperature in the space, so to speak. More work in this area means more capability, more complexity, more risk of misuse, and a heightened risk of the possibility of additional actors. Look no further than the theory of deterrence, and how it shifted from a tolerable short term strategy during the Cold War (one with few actors), to a high risk strategy as the trove of technology was enlarged, the likelihood of it spreading increased, and the number of active and potential actors increased dramatically also.

In my view AI weapons development demands a judicious, steady, paced, and thoroughly informed approach to keep the whole space from burning up —- figuratively and literally.