The strategy's focus on AI simulation acknowledges a key risk: AI systems can develop winning tactics by exploiting unrealistic aspects of a simulation. If simulation physics or capabilities don't perfectly match reality, these AI-derived strategies could fail catastrophically when deployed.

While the U.S. and China pursue hyperwar as a national strategy, its most rapid development is happening organically on the battlefield. Outnumbered forces like Ukraine are forced to innovate with autonomous systems out of necessity, driving a bottom-up adoption of hyperwar tactics.

To test and train AI pilots, Shield AI acquired simulation leader Echelon. This is critical because physical training ranges are too small and limited to rehearse for vast, complex theaters like the Pacific. High-fidelity simulation becomes the only way to develop and validate autonomy at scale.

Instead of automating decisions, the Pentagon's AI strategy focuses on synthesizing vast amounts of data—assets, weather, potential reactions—to expand a human operator's situational awareness, enabling them to make better, more informed choices.

Grant Demaree estimates that achieving "perfect" military decisions through AI software—affecting everything from kill chains to logistics and planning—would multiply U.S. combat power by eightfold. This massive leverage makes it the most critical area for ensuring national security and global stability.

Unlike nuclear weapons, which don't create better versions of themselves, AI systems can improve their own capabilities. This creates a recursive loop where the first entity to achieve a breakthrough gains a runaway intelligence advantage, dominating all rivals technologically and militarily.

The most effective AI architecture for complex tasks involves a division of labor. An LLM handles high-level strategic reasoning and goal setting, providing its intent in natural language. Specialized, efficient algorithms then translate that strategic intent into concrete, tactical actions.

Since AI agents dramatically lower the cost of building solutions, the premium on getting it perfect the first time diminishes. The new competitive advantage lies in quickly launching and iterating on multiple solutions based on real-world outcomes, rather than engaging in exhaustive upfront planning.

Smack Technologies argues that general-purpose LLMs fail in military strategy because they rely on historical labeled data. For novel, high-stakes conflicts, a different approach like deep reinforcement learning is required, training models within physics-grounded simulations of potential future battlefields.