More than a decade of disparate but intense international corporate effort has been lavished on the development of a wide variety of driverless car technologies. The failure of that effort, beyond the most modest of roadway topologies and dynamics, carries serious implications for success of AI systems on Clausewitz landscapes of fog, friction, and skilled adversarial agency. Matters of AI-driven control system stability and suitability are subtle, technically complex, and relentlessly challenging. Political and military discussions of these issues have been blind-sided by self-interested salesmen of the technology and are dangerously simplistic. A central threat is that, after considerable further investment, AI-driven weapons and related systems will, like driverless cars confronting chaotic, rapidly-changing environments, simply not work reliably in real combat, causing not only mission failure, but massive levels of civilian ‘collateral damage’ and consequent blowback.                              — American Automobile Association (2022)                                  — Col. Guilong Yan (2020)                                  — Dr. D.N. Wallace (1981)