Boyd and the Machine: Teach Warfighters to Master AI

U.S. Naval Institute | Julie Chase

National Security Presidential Memorandum (NSPM-11), issued by President Donald Trump on 5 June, mandates that the joint force develop and implement an artificial intelligence curriculum by an October deadline. This directive forces the military to rapidly train personnel to employ generative AI systems while maintaining critical thinking skills on the high-stakes national security frontier.

To prevent operators from outsourcing cognitive labor and triggering algorithmic hallucinations, a proposed 'Context Block' framework bounds these tools within Colonel John Boyd’s OODA Loop. This platform-agnostic cognitive architecture requires warfighters to inject their tacit knowledge and commander's intent during the orientation phase before the machine generates outputs. AI lacks human judgment. Consequently, the framework acts as a localized, last-mile safeguard, complementing enterprise-level guardrails. Grassroots validation occurred in April 2026 at the Modern Day Marine AI Hackathon, where the 'Semper AI' team successfully prototyped the interface using a simulated logistics scenario.

Comment

Integrating generative artificial intelligence into the joint force under National Security Presidential Memorandum 11 (NSPM-11) exposes a fundamental tension in modern military decision-making. By bounding algorithmic outputs within Colonel John Boyd’s OODA Loop, the proposed Context Block framework attempts to preserve human cognitive supremacy during the critical orientation phase. This approach prevents the delegation of tactical judgement to large language models that lack moral agency.

The downstream consequence of this doctrinal shift will be felt in the design of future command post exercises, such as those conducted at Marine Corps Base Quantico. Staff officers will require rigorous training to construct standardised prompt chains rather than relying on intuitive, unstructured queries. Ultimately, the operational utility of the Context Block depends on whether the U.S. Marine Corps can codify these cognitive boundaries into formal tactical publications before adversaries exploit unconstrained algorithmic hallucinations.

Strategic Question for Discussion

If the U.S. Marine Corps successfully integrates the Context Block into tactical workflows, does this framework mitigate the risk of algorithmic hallucinations, or does it merely introduce a new point of cognitive failure during high-tempo operations?

The trajectory indicates that while the Context Block provides a necessary cognitive speed bump, its reliance on manual human input may fail under the extreme time compression of peer-adversary electronic warfare. My assessment is that the framework will successfully prevent hallucinations in deliberate planning phases, but will likely be bypassed by operators during active, high-tempo engagements where automated decision-support systems offer faster, albeit riskier, solutions.

Share your assessment in the comments below.

To read in details click here